ci: merge test+build+push into single pipeline (#1)

Merge ci.yaml + build-push.yml into single CI pipeline. Single job: vet → test → build binary → build image → push. Image push gated on main push only. Fixed Dockerfile to golang:1.26, build cmd/worker, removed HTTP healthcheck.

---------

Co-authored-by: Test <[email protected]>
Reviewed-on: #1

.dockerignore

@@ -0,0 +1,11 @@
+.git
+.gitignore
+*.md
+.env.local
+.env
+tests/
+*.test.go
+coverage/
+.DS_Store
+k8s/
+migrations/

.gitea/workflows/ci.yaml

@@ -1,4 +1,4 @@
-name: ci
+name: CI
 
 on:
   push:
@@ -6,14 +6,14 @@ on:
   pull_request:
 
 jobs:
-  test:
+  test-build-push:
     runs-on: golang
     container:
-      image: golang:1.25
+      image: golang:1.26
     env:
       GOPRIVATE: forgejo.riotpiao.com
-      GOFLAGS: -mod=readonly
-      GITHUB_TOKEN: ${{ secrets.REGISTRY_PAT }}
+      REGISTRY: forgejo.riotpiao.com
+      IMAGE: forgejo.riotpiao.com/rock/poimen-workflows
     steps:
       - name: Configure git authentication
         run: |
@@ -25,17 +25,38 @@ jobs:
         run: |
           git init
           git remote add origin https://forgejo.riotpiao.com/rock/poimen-workflows.git
-          git fetch origin ${{ github.ref_name }} --depth=1
+          git fetch origin ${{ github.head_ref || github.ref_name }} --depth=1
           git checkout FETCH_HEAD
 
       - name: Download dependencies
         run: go mod download
 
-      - name: Test
-        run: go test -v ./...
-      
-      - name: Build
-        run: go build -o /tmp/poimen-bin/ ./cmd/...
-      
       - name: Vet
         run: go vet ./...
+
+      - name: Test
+        run: go test ./...
+
+      - name: Build binary
+        run: CGO_ENABLED=0 GOOS=linux go build -o /tmp/poimen-worker ./cmd/worker
+
+      - name: Get short SHA
+        id: sha
+        run: echo "short_sha=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+
+      - name: Registry login
+        if: github.ref == 'refs/heads/main' && github.event_name == 'push'
+        run: |
+          echo "${{ secrets.FORGEJO_REGISTRY_TOKEN }}" | docker login "${REGISTRY}" \
+            --username "${{ secrets.FORGEJO_REGISTRY_USER }}" --password-stdin
+
+      - name: Build and push image
+        if: github.ref == 'refs/heads/main' && github.event_name == 'push'
+        run: |
+          docker build \
+            -t "${IMAGE}:${{ steps.sha.outputs.short_sha }}" \
+            -t "${IMAGE}:latest" \
+            .
+          docker push "${IMAGE}:${{ steps.sha.outputs.short_sha }}"
+          docker push "${IMAGE}:latest"
+          echo "✓ Pushed ${IMAGE}:${{ steps.sha.outputs.short_sha }}"

.sops.yaml

@@ -0,0 +1,3 @@
+creation_rules:
+  - path_regex: k8s/.*\.enc\.ya?ml
+    age: age1e5fq3hwxy78psus2nfvmtmua36g0u3suk78ephw6246l974d2utsvn0hla

DEPLOYMENT.md

@@ -0,0 +1,293 @@
+# Poimen Application Deployment
+
+## Overview
+
+Poimen is a unified application consisting of three services:
+- **poimen-memory**: Memory/Graph RAG service
+- **poimen-workflows**: Temporal orchestration + API
+- **poimen-frontend**: Next.js frontend
+
+All services are deployed together as a single application in the `poimen` namespace.
+
+## Local Development
+
+### Prerequisites
+- Docker
+- Docker Compose
+- Node.js 18+
+- Go 1.21+
+- Python 3.11+
+
+### Start Local Stack
+
+```bash
+docker-compose up -d
+```
+
+This starts:
+- PostgreSQL (memory + workflows DBs)
+- Redis (cache)
+- Temporal (workflow orchestration)
+- poimen-memory (8000)
+- poimen-workflows (8080)
+- poimen-workflows-worker
+- poimen-frontend (3000)
+
+### Access Services
+
+- Frontend: http://localhost:3000
+- Workflows API: http://localhost:8080
+- Memory API: http://localhost:8000
+- Temporal UI: http://localhost:8233
+
+### Stop Stack
+
+```bash
+docker-compose down
+```
+
+## Building & Pushing Images
+
+### Build All Services
+
+```bash
+./build-push.sh latest
+```
+
+Or specific services:
+
+```bash
+docker build -t forgejo.riotpiao.com/rock/poimen-memory:v1.0.0 ./memory
+docker push forgejo.riotpiao.com/rock/poimen-memory:v1.0.0
+```
+
+### Image Tagging Strategy
+
+- `latest`: Development/staging
+- `v1.0.0`, `v1.0.1`, etc.: Production releases
+- `main-{commit-hash}`: CI/CD automated builds
+
+## Kubernetes Deployment
+
+### Prerequisites
+
+- Kubernetes cluster (1.24+)
+- kubectl configured
+- Kustomize installed
+- Registry credentials configured
+
+### Deploy to Cluster
+
+```bash
+cd k8s
+./deploy.sh -a
+```
+
+Or with specific tags:
+
+```bash
+./deploy.sh -m v1.0.0 -w v1.0.0 -f v1.0.0
+```
+
+### Verify Deployment
+
+```bash
+kubectl get pods -n poimen
+kubectl get svc -n poimen
+kubectl logs -n poimen -l app=poimen-workflows
+```
+
+## Configuration
+
+### Environment Variables
+
+Configure in `k8s/poimen-application.yaml` under `spec.template.spec.env`:
+
+**Common:**
+- `TEMPORAL_HOST`: Temporal server (default: temporal:7233)
+- `DATABASE_URL`: PostgreSQL connection
+- `JWT_SECRET`: JWT signing key
+- `LOG_LEVEL`: debug|info|warn|error
+
+**Memory Service:**
+- `REDIS_URL`: Redis connection
+- `ELASTICSEARCH_URL`: Optional full-text search
+
+**Workflows Service:**
+- `MEMORY_SERVICE_URL`: Internal memory service URL
+
+**Frontend:**
+- `NEXT_PUBLIC_WORKFLOWS_API`: External workflows API
+- `NEXT_PUBLIC_MEMORY_API`: External memory API
+- `OAUTH_CLIENT_ID`, `OAUTH_CLIENT_SECRET`: Auth provider
+
+### Secrets
+
+Create secrets before deployment:
+
+```bash
+kubectl create secret generic poimen-db-credentials \
+  --from-literal=memory-url="postgresql://..." \
+  --from-literal=workflows-url="postgresql://..." \
+  -n poimen
+
+kubectl create secret generic poimen-secrets \
+  --from-literal=jwt-secret="..." \
+  --from-literal=oauth-client-id="..." \
+  --from-literal=oauth-client-secret="..." \
+  -n poimen
+```
+
+## Architecture
+
+```
+┌─────────────────────────────────────────────┐
+│             LoadBalancer Service            │
+│          poimen-frontend:80→3000            │
+└─────────────────┬───────────────────────────┘
+                  │
+          ┌───────┴────────┐
+          ▼                ▼
+   ┌──────────────┐  ┌──────────────┐
+   │   Frontend   │  │  Workflows   │
+   │   (3000)     │  │   API (8080) │
+   │   2 replicas │  │  2 replicas  │
+   └──────────────┘  └──────┬───────┘
+          │                 │
+          │           ┌─────┴─────┐
+          │           ▼           ▼
+          │      ┌─────────────────────┐
+          │      │  Temporal Cluster   │
+          │      │  (External)         │
+          │      └─────────────────────┘
+          │
+          └──────────────────┬──────────────┐
+                             ▼              ▼
+                    ┌──────────────┐  ┌──────────────┐
+                    │   Memory     │  │  PostgreSQL  │
+                    │  (8000)      │  │  (5432)      │
+                    │  1 replica   │  │              │
+                    └──────────────┘  └──────────────┘
+```
+
+## Scaling
+
+### Horizontal Scaling
+
+Adjust replicas in `k8s/poimen-application.yaml`:
+
+```yaml
+spec:
+  replicas: 3  # Increase this
+```
+
+Or patch:
+
+```bash
+kubectl patch deployment poimen-workflows -p '{"spec":{"replicas":3}}' -n poimen
+```
+
+### Resource Requests/Limits
+
+Add to container spec:
+
+```yaml
+resources:
+  requests:
+    cpu: 100m
+    memory: 256Mi
+  limits:
+    cpu: 500m
+    memory: 512Mi
+```
+
+## Monitoring & Logging
+
+### Check Status
+
+```bash
+kubectl get pods -n poimen -w
+kubectl describe pod <pod-name> -n poimen
+kubectl logs -n poimen -f -l app=poimen-workflows --all-containers=true
+```
+
+### Health Checks
+
+All services expose `/health` endpoint:
+
+```bash
+curl http://poimen-workflows:8080/health
+curl http://poimen-memory:8000/health
+curl http://poimen-frontend:3000/
+```
+
+## Updates & Rollbacks
+
+### Rolling Update
+
+```bash
+./deploy.sh -w v1.0.1
+```
+
+Kubernetes automatically rolls out with health checks.
+
+### View Rollout Status
+
+```bash
+kubectl rollout status deploy/poimen-workflows -n poimen
+```
+
+### Rollback
+
+```bash
+kubectl rollout undo deploy/poimen-workflows -n poimen
+```
+
+## Troubleshooting
+
+### Services Can't Connect
+
+Check service DNS:
+```bash
+kubectl run -it --rm debug --image=busybox --restart=Never -- nslookup poimen-workflows
+```
+
+### Database Migrations Failing
+
+```bash
+kubectl exec -it <workflows-pod> -n poimen -- \
+  ./workflows migrate up
+```
+
+### Temporal Worker Not Picking Up Activities
+
+Check worker logs:
+```bash
+kubectl logs -n poimen -l app=poimen-workflows --all-containers=true | grep -i activity
+```
+
+Verify activities registered in `cmd/worker/main.go`
+
+## CI/CD Integration
+
+### GitHub Actions Example
+
+```yaml
+name: Build & Push Poimen
+
+on:
+  push:
+    branches: [main]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - name: Build & Push
+        run: ./build-push.sh main-${{ github.sha }}
+```
+
+### Automatic Deployment
+
+Configure ArgoCD to watch `k8s/` directory for updates.

Dockerfile

@@ -1,385 +1,20 @@
-# Multi-stage build for Poimen Temporal Worker
-# Stage 1: Builder - Compile Go binary and set up tools
-FROM golang:1.25-alpine AS builder
-
-WORKDIR /build
-
-# Install system dependencies (ast-grep, git, build essentials)
-RUN apk add --no-cache \
-    git \
-    curl \
-    wget \
-    ca-certificates \
-    gcc \
-    musl-dev \
-    bash \
-    && echo "[builder] System dependencies installed"
-
-# Install ast-grep CLI tool
-RUN curl -fsSL https://github.com/ast-grep/ast-grep/releases/download/0.24.0/sg-x86_64-unknown-linux-musl.tar.gz \
-    | tar xzf - -C /usr/local/bin \
-    && chmod +x /usr/local/bin/sg \
-    && sg --version \
-    && echo "[builder] ast-grep installed"
-
-# Install Node.js for pi CLI and browser-use
-RUN apk add --no-cache nodejs npm \
-    && echo "[builder] Node.js installed"
-
-# Install pi CLI globally
-RUN npm install -g @earendil-works/pi-coding-agent --unsafe-perm \
-    && pi --version \
-    && echo "[builder] pi CLI installed"
-
-# Install browser-use CLI for browser automation
-RUN npm install -g browser-use --unsafe-perm \
-    && browser-use --version \
-    && echo "[builder] browser-use CLI installed"
-
-# Set up pi home directory and skills
-RUN mkdir -p ~/.pi/agent/skills ~/.pi/agent/agents \
-    && echo "[builder] pi directories created"
-
-# Stage 2: Download pi skills (caveman & andrej karpathy)
-# Clone caveman skill from pi-agent repo
-RUN cd /tmp && git clone https://github.com/earendil-works/pi-agent.git pi-repo \
-    && mkdir -p ~/.pi/agent/skills/caveman \
-    && cp -r pi-repo/examples/skills/caveman/* ~/.pi/agent/skills/caveman/ 2>/dev/null || true \
-    && echo "[builder] caveman skill installed"
-
-# Create andrej karpathy skill manually (reference/training patterns)
-RUN mkdir -p ~/.pi/agent/skills/andrej-karpathy && cat > ~/.pi/agent/skills/andrej-karpathy/SKILL.md << 'EOF'
-# Andrej Karpathy LLM & AI Principles Skill
-
-Build neural networks and LLM systems with proven patterns from Andrej Karpathy.
-Topics: attention mechanisms, transformer training, inference optimization, edge cases.
-
-## Key Principles
-
-### 1. Simplicity First
-- Start with minimal implementation
-- Add complexity only when justified
-- Test each component independently
-- Use debugging tools effectively
-
-### 2. Neural Network Architecture
-- Understand backward pass deeply
-- Implement from scratch when possible
-- Use visualization for debugging
-- Profile before optimizing
-
-### 3. LLM Training Patterns
-- Quality data > quantity
-- Curriculum learning for complex tasks
-- Loss landscape visualization
-- Checkpoint strategy matters
-
-### 4. Inference Optimization
-- Quantization without quality loss
-- KV cache management
-- Batch processing strategies
-- Latency profiling
-
-### 5. Failure Analysis
-- Log intermediate activations
-- Check gradient flow
-- Validate data pipeline
-- Test edge cases explicitly
-
-## Usage in Poimen
-
-Apply when:
-- Designing workflow stages (like training curricula)
-- Optimizing inference (planner/judge/implementer prompts)
-- Debugging convergence issues (retry patterns)
-- Scaling to production (quantization patterns)
-
-## Resources
-- github.com/karpathy/minGPT - Minimal GPT implementation
-- youtube: "Neural Networks: Zero to Hero" series
-- Papers: Attention Is All You Need, GPT series whitepapers
-EOF
-    && echo "[builder] andrej-karpathy skill created"
-
-# Create browser-use skill for web testing & automation
-RUN mkdir -p ~/.pi/agent/skills/browser-use && cat > ~/.pi/agent/skills/browser-use/SKILL.md << 'EOF'
-# browser-use: Browser Automation Skill
-
-Automate web browser interactions for testing, verification, and UI validation.
-Topics: headless browser control, visual testing, form automation, screenshot capture.
-
-## Key Capabilities
-
-### 1. Browser Control
-- Launch headless Chrome/Firefox
-- Navigate to URLs
-- Wait for elements/navigation
-- Handle popups/dialogs
-
-### 2. Interaction Patterns
-- Click buttons/links
-- Fill forms (text, dropdown, checkbox)
-- Drag & drop
-- Keyboard input
-
-### 3. Verification & Capture
-- Screenshot capture
-- Element inspection
-- Accessibility checks
-- Network monitoring
-
-### 4. Wait Strategies
-- Wait for element visible
-- Wait for navigation
-- Wait for condition (custom JS)
-- Timeout handling
-
-### 5. Error Recovery
-- Retry failed actions
-- Handle stale elements
-- Browser crash recovery
-- Memory leak prevention
-
-## Usage in Poimen Phases
-
-### Phase T2 (Implementation)
-- Test generated UI code in real browser
-- Verify visual layout matches spec
-- Validate form inputs work correctly
-
-### Phase T3 (Verification)
-- Visual regression testing
-- Accessibility validation (ARIA, keyboard nav)
-- Cross-browser verification
-
-### Phase T6 (Integration)
-- End-to-end workflow testing
-- External service integration testing
-- User journey verification
-
-### Phase T9 (Release)
-- Pre-release smoke tests
-- Deployment verification
-- Production canary testing
-
-## Example Workflows
-
-```bash
-# Launch browser and take screenshot
-browser-use screenshot "https://example.com" --file output.png
-
-# Fill form and submit
-browser-use interact "https://example.com" \
-  --click "#submit-btn" \
-  --type "#email" "[email protected]" \
-  --type "#password" "secretpass" \
-  --click ".submit"
-
-# Wait for dynamic content and extract data
-browser-use extract "https://example.com" \
-  --wait ".dynamic-content" \
-  --selector ".data-row" \
-  --output json
-
-# Accessibility audit
-browser-use audit "https://example.com" \
-  --check wcag2a \
-  --report a11y-report.html
-```
-
-## Integration with Poimen
-
-Pre-generated code can be tested:
-```bash
-# Generate code (T2)
-implementer_output = "function handleClick() { ... }"
-
-# Verify in browser (T3)
-browser-use interact "http://localhost:3000" \
-  --click ".test-button" \
-  --screenshot result.png
-
-# Compare with expected
-verify_visual_match(result.png, expected.png)
-```
-
-## Performance Notes
-
-- Startup: ~2-5s per browser
-- Action latency: 100-500ms per interaction
-- Screenshot: 500ms-2s (depends on page size)
-- Keep browser alive for batch operations (pool management)
-
-## Error Handling
-
-- Transient: Network timeout → retry with backoff
-- Permanent: Element not found → fail and log
-- Flaky: Wait strategies → increase timeout gradually
-- Memory: Reuse browser instances → kill after 10 uses
-
-## Resources
-- docs.browseruse.com - Official documentation
-- github.com/browser-use/browser-use - Source code
-- Chrome DevTools Protocol - Advanced browser control
-EOF
-    && echo "[builder] browser-use skill created"
-
-# Copy Go source code
-COPY . /build/
-
-# Download Go dependencies
-RUN go mod download \
-    && echo "[builder] Go dependencies downloaded"
-
-# Build worker binary
-RUN CGO_ENABLED=1 GOOS=linux go build -o /build/worker ./cmd/worker \
-    && echo "[builder] Worker binary built"
-
-# Verify binary
-RUN file /build/worker && ls -lh /build/worker
-
-# Stage 3: Runtime - Minimal base image with runtime dependencies
-FROM alpine:3.20
-
-LABEL maintainer="Poimen Team"
-LABEL description="Poimen Temporal Worker with memory service, ast-grep, and browser automation"
+FROM golang:1.26-alpine AS builder
 
 WORKDIR /app
+COPY go.mod go.sum ./
+RUN go mod download
+COPY . .
 
-# Install runtime dependencies (including Chromium for browser-use)
-RUN apk add --no-cache \
-    ca-certificates \
-    git \
-    bash \
-    curl \
-    jq \
-    chromium \
-    chromium-chromedriver \
-    && echo "[runtime] Runtime dependencies installed"
-
-# Install Node.js for pi CLI and browser-use
-RUN apk add --no-cache nodejs npm \
-    && echo "[runtime] Node.js installed"
-
-# Install pi CLI in runtime image
-RUN npm install -g @earendil-works/pi-coding-agent --unsafe-perm \
-    && pi --version \
-    && echo "[runtime] pi CLI installed"
-
-# Install browser-use CLI in runtime image
-RUN npm install -g browser-use --unsafe-perm \
-    && browser-use --version \
-    && echo "[runtime] browser-use CLI installed"
-
-# Copy ast-grep binary from builder
-COPY --from=builder /usr/local/bin/sg /usr/local/bin/sg
-RUN chmod +x /usr/local/bin/sg && sg --version \
-    && echo "[runtime] ast-grep copied"
-
-# Copy pi skills from builder
-COPY --from=builder /root/.pi /root/.pi
-RUN ls -la /root/.pi/agent/skills/ \
-    && echo "[runtime] pi skills configured"
-
-# Copy worker binary from builder
-COPY --from=builder /build/worker /app/worker
-RUN chmod +x /app/worker && file /app/worker \
-    && echo "[runtime] Worker binary copied"
-
-# Create app directory structure
-RUN mkdir -p /app/work /app/logs /app/screenshots \
-    && chmod 755 /app/work /app/logs /app/screenshots \
-    && echo "[runtime] App directories created"
-
-# Health check endpoint
-EXPOSE 8081
-
-# Worker task queue listener
-ENV TEMPORAL_NAMESPACE=poimen-harness \
-    TEMPORAL_HOSTPORT=temporal-frontend.temporal:7233 \
-    MEMORY_SERVICE_URL=http://memory-service.poimen:5000 \
-    MEMORY_SERVICE_TOKEN= \
-    ANTHROPIC_API_KEY= \
-    PI_SKILLS_PATH=/root/.pi/agent/skills \
-    AST_GREP_BIN=/usr/local/bin/sg \
-    BROWSER_USE_BIN=/usr/local/bin/browser-use \
-    CHROMIUM_BIN=/usr/bin/chromium-browser \
-    SCREENSHOTS_DIR=/app/screenshots
-
-# Entrypoint script with startup diagnostics
-COPY --chmod=755 << 'EOF' /app/entrypoint.sh
-#!/bin/bash
-set -e
-
-echo "[$(date)] ========== POIMEN WORKER STARTUP =========="
-echo "[$(date)] Container: $HOSTNAME"
-echo "[$(date)] Image: $(cat /etc/os-release | grep PRETTY_NAME | cut -d= -f2)"
-
-# Verify CLI tools
-echo "[$(date)] ✓ Checking CLI tools..."
-echo "  - Go version: $(go version 2>/dev/null || echo 'N/A')"
-echo "  - ast-grep: $(sg --version 2>&1 | head -1)"
-echo "  - pi: $(pi --version 2>&1 | head -1)"
-echo "  - browser-use: $(browser-use --version 2>&1 | head -1)"
-echo "  - chromium: $(chromium-browser --version 2>&1 || echo 'Not found')"
-echo "  - git: $(git --version)"
-echo "  - node: $(node --version)"
-echo "  - npm: $(npm --version)"
-
-# Verify pi skills
-echo "[$(date)] ✓ Checking pi skills..."
-if [ -d "$PI_SKILLS_PATH" ]; then
-    echo "  - Skills path: $PI_SKILLS_PATH"
-    ls -1 "$PI_SKILLS_PATH" | sed 's/^/    ✓ /'
-else
-    echo "  - WARNING: Skills path not found: $PI_SKILLS_PATH"
-fi
-
-# Verify browser tools
-echo "[$(date)] ✓ Checking browser automation tools..."
-echo "  - Chromium binary: $CHROMIUM_BIN"
-echo "  - Screenshots directory: $SCREENSHOTS_DIR"
-if [ -d "$SCREENSHOTS_DIR" ]; then
-    echo "  - Screenshots dir ready ($(du -sh $SCREENSHOTS_DIR 2>/dev/null | cut -f1 || echo '0B'))"
-fi
-
-# Check environment variables
-echo "[$(date)] ✓ Configuration loaded:"
-echo "  - TEMPORAL_NAMESPACE: $TEMPORAL_NAMESPACE"
-echo "  - TEMPORAL_HOSTPORT: $TEMPORAL_HOSTPORT"
-echo "  - MEMORY_SERVICE_URL: ${MEMORY_SERVICE_URL:-(not set)}"
-echo "  - PI_SKILLS_PATH: $PI_SKILLS_PATH"
-echo "  - CHROMIUM_BIN: $CHROMIUM_BIN"
-
-# Verify memory service connectivity (optional, non-blocking)
-if [ ! -z "$MEMORY_SERVICE_URL" ]; then
-    echo "[$(date)] ✓ Testing memory service connectivity..."
-    if curl -sf "$MEMORY_SERVICE_URL/health" > /dev/null 2>&1; then
-        echo "  - Memory service: HEALTHY"
-    else
-        echo "  - Memory service: UNREACHABLE (will retry in worker)"
+# Re-use CI-built binary if present, otherwise build
+ARG BINARY_PATH=
+RUN if [ -n "$BINARY_PATH" ] && [ -f "$BINARY_PATH" ]; then \
+      cp "$BINARY_PATH" worker; \
+    else \
+      CGO_ENABLED=0 GOOS=linux go build -o worker ./cmd/worker; \
     fi
-fi
 
-# Test browser automation (optional, non-blocking)
-echo "[$(date)] ✓ Testing browser automation..."
-if command -v chromium-browser &> /dev/null && command -v browser-use &> /dev/null; then
-    echo "  - Chromium available: YES"
-    echo "  - browser-use available: YES"
-    echo "  - Browser automation: READY"
-else
-    echo "  - Browser automation: WARNING - missing dependencies"
-fi
-
-echo "[$(date)] ========== STARTING WORKER =========="
-exec /app/worker
-EOF
-
-RUN chmod +x /app/entrypoint.sh
-
-# Run worker with diagnostics
-ENTRYPOINT ["/app/entrypoint.sh"]
-
-# Health check
-HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
-    CMD curl -f http://localhost:8081/health || exit 1
+FROM alpine:3.20
+RUN apk --no-cache add ca-certificates
+WORKDIR /app
+COPY --from=builder /app/worker .
+ENTRYPOINT ["./worker"]

README.md

@@ -120,6 +120,7 @@ Poimen embodies three core principles:
 
 | Activity | Purpose | Timeout | Retry |
 |----------|---------|---------|-------|
+| **AssumeRoleActivity** | Request temporary JWT token (like AWS AssumeRole) | 30s | 2x |
 | **CloneRepo** | Clone git repository | 30s | 3x |
 | **AnalyzeCode** | Static analysis (SAST) | 120s | 2x |
 | **SecurityScan** | Dependency & vulnerability scan | 60s | 2x |
@@ -546,10 +547,8 @@ Poimen supports multiple authentication methods to federate LLM access across cu
 #### 1. Bearer Token (JWT/OAuth2)
 ```go
 auth := &routing.LLMAuth{
-    Type:     routing.AuthTypeBearer,
-    Token:    "eyJhbGciOiJIUzI1NiIs...",  // JWT token from your auth provider
-    TenantID: "customer-123",               // Optional: tenant isolation
-    Scopes:   "llm:read llm:write",        // Optional: OAuth2 scopes
+    Type:  routing.AuthTypeBearer,
+    Token: "eyJhbGciOiJIUzI1NiIs...",  // JWT token from your auth provider
 }
 client := routing.NewLLMClientWithAuth(auth)
 ```
@@ -557,9 +556,8 @@ client := routing.NewLLMClientWithAuth(auth)
 #### 2. API Key
 ```go
 auth := &routing.LLMAuth{
-    Type:     routing.AuthTypeAPIKey,
-    APIKey:   "sk-xxx-yyy-zzz",  // API key from provider
-    TenantID: "customer-456",
+    Type:   routing.AuthTypeAPIKey,
+    APIKey: "sk-xxx-yyy-zzz",  // API key from provider
 }
 client := routing.NewLLMClientWithAuth(auth)
 ```
@@ -570,7 +568,6 @@ auth := &routing.LLMAuth{
     Type:        routing.AuthTypeCustom,
     HeaderName:  "X-Custom-Auth",
     HeaderValue: "custom-token-value",
-    TenantID:    "customer-789",
 }
 client := routing.NewLLMClientWithAuth(auth)
 ```
@@ -581,49 +578,41 @@ config := &routing.LLMRouterConfig{
     Provider:      openaiProvider,
     KnowledgeBase: kb,
     Auth: &routing.LLMAuth{
-        Type:     routing.AuthTypeBearer,
-        Token:    jwtToken,
-        TenantID: customerID,  // Tenant isolation in multi-tenant deployments
+        Type:  routing.AuthTypeBearer,
+        Token: jwtToken,
     },
-    TenantID: customerID,  // Additional tenant tracking
 }
 router := routing.NewLLMRouter(config)
 ```
 
-#### Multi-Tenant Federated Access
+#### Per-Deployment Auth
 
-For multi-tenant deployments:
+Each Poimen deployment gets its own LLM token:
 
 ```go
-// Per-customer isolated routers
-func CreateCustomerRouter(customerID, jwtToken string) (*routing.LLMRouter, error) {
+// In K8s secret/vault
+LLM_AUTH_TOKEN="eyJhbGciOiJIUzI1NiIs..."
+
+// In code
+func InitializeRouter() (*routing.LLMRouter, error) {
+    token := os.Getenv("LLM_AUTH_TOKEN")
+    if token == "" {
+        return nil, fmt.Errorf("LLM_AUTH_TOKEN not set")
+    }
+    
     auth := &routing.LLMAuth{
-        Type:     routing.AuthTypeBearer,
-        Token:    jwtToken,
-        TenantID: customerID,  // Passed to LLM API in X-Tenant-ID header
-        Scopes:   "llm:read",  // Restrict scopes per customer
+        Type:  routing.AuthTypeBearer,
+        Token: token,
     }
     
     config := &routing.LLMRouterConfig{
         Provider:      &routing.LLMClient{},
-        KnowledgeBase: globalKB,
+        KnowledgeBase: kb,
         Auth:          auth,
-        TenantID:      customerID,
     }
     
     return routing.NewLLMRouter(config)
 }
-
-// Usage in activity
-input := &routing.LLMRouterInput{
-    Message:   "Analyze code",
-    MemoryContext: memCtx,
-}
-output, err := router.Route(ctx, input)
-// Headers sent to LLM API:
-// Authorization: Bearer <jwtToken>
-// X-Tenant-ID: <customerID>
-// X-OAuth-Scopes: llm:read
 ```
 
 #### Token Refresh & Rotation
@@ -649,10 +638,8 @@ result, err := client.Chat(ctx, systemPrompt, userMsg)
 
 | Header | Set When | Value | Purpose |
 |--------|----------|-------|---------|
-| `Authorization` | Bearer auth | `Bearer {token}` | OAuth2/JWT authentication |
+| `Authorization` | Bearer auth | `Bearer {token}` | JWT/OAuth2 authentication |
 | `X-API-Key` | API Key auth | `{api-key}` | API key authentication |
-| `X-Tenant-ID` | Any auth type | `{tenantID}` | Tenant/customer isolation |
-| `X-OAuth-Scopes` | Bearer with scopes | Space-separated scopes | OAuth2 scope enforcement |
 | Custom header | Custom auth | `{headerValue}` | Custom authentication scheme |
 
 ### LLMRouter Configuration (Programmatic)
@@ -780,95 +767,171 @@ kubectl apply -f k8s/worker-deployment.yaml
 
 ---
 
-## Security & Multi-Tenancy
+## Security & Token Management
 
-### Token Management
+### Never Hardcode Tokens
 
-**Never hardcode tokens!** Use secure secret management:
+**Use secure secret management:**
 
 ```go
 // ❌ DON'T DO THIS
 auth := &routing.LLMAuth{
+    Type:  routing.AuthTypeBearer,
     Token: "eyJhbGciOiJIUzI1NiIs...",  // Hardcoded!
 }
 
 // ✅ DO THIS
-tokenFromVault, _ := vaultClient.GetSecret("llm-token-" + customerID)
+token := os.Getenv("LLM_AUTH_TOKEN")
 auth := &routing.LLMAuth{
-    Type:     routing.AuthTypeBearer,
-    Token:    tokenFromVault,
-    TenantID: customerID,
+    Type:  routing.AuthTypeBearer,
+    Token: token,
 }
 ```
 
 **Recommended secret management:**
-- Kubernetes Secrets (development)
-- HashiCorp Vault (production)
-- AWS Secrets Manager / GCP Secret Manager (cloud)
-- Sealed Secrets / Sealed Policies
+- **Kubernetes Secrets** (development) — stored in etcd
+- **HashiCorp Vault** (production) — centralized secret management
+- **AWS Secrets Manager** (cloud) — managed service
+- **GCP Secret Manager** (cloud) — managed service
+- **Sealed Secrets** or **Sealed Policies** — encrypted in git
 
-### Tenant Isolation
+### Token Rotation & Refresh
 
-The `TenantID` header enforces tenant isolation on the LLM API side:
+For long-running Temporal workflows, refresh tokens before they expire:
 
 ```go
-// Customer A's workflow
-authA := &routing.LLMAuth{
-    Type:     routing.AuthTypeBearer,
-    Token:    tokenA,
-    TenantID: "customer-A",  // ← Isolates this customer
-}
+client := routing.NewLLMClientWithAuth(auth)
 
-// Customer B's workflow
-authB := &routing.LLMAuth{
-    Type:     routing.AuthTypeBearer,
-    Token:    tokenB,
-    TenantID: "customer-B",  // ← Isolates this customer
+// Later: token expires
+newToken := os.Getenv("LLM_AUTH_TOKEN_REFRESHED")
+newAuth := &routing.LLMAuth{
+    Type:  routing.AuthTypeBearer,
+    Token: newToken,
 }
+client.UpdateAuth(newAuth)
 ```
 
-The LLM API server should:
-- Validate tenant ownership of tokens
-- Enforce data boundaries per tenant
-- Log access per tenant ID
-- Rate-limit per tenant
+### Authorization: LLM API Side
 
-### Scope-Based Access Control
+The LLM provider (api.riotpiao.com) should:
+- Validate JWT signature & expiration
+- Enforce API rate limits per token
+- Log all requests with token identity
+- Support token revocation / blacklisting
 
-Use OAuth2 scopes to limit capabilities:
+---
+
+## AssumeRoleActivity: Temporary LLM Token Grants
+
+**Like AWS AssumeRole**, AssumeRoleActivity requests temporary credentials for accessing LLM APIs:
 
 ```go
-// Analytics-only customer
-analytics := &routing.LLMAuth{
-    Type:   routing.AuthTypeBearer,
-    Token:  token,
-    Scopes: "llm:read",  // Read-only
+// 1. User requests temporary token
+assumeRoleInput := &routing.AssumeRoleInput{
+    Identity: "[email protected]",           // Who is accessing
+    Scope:    "llm:read llm:write",        // What permissions
+    DurationSeconds: 1800,                 // 30 minutes
 }
 
-// Full-access customer
-admin := &routing.LLMAuth{
-    Type:   routing.AuthTypeBearer,
-    Token:  token,
-    Scopes: "llm:read llm:write llm:admin",  // Full access
+// 2. Activity exchanges with auth server → returns JWT
+output, err := temporalClient.ExecuteActivity(ctx, 
+    routing.AssumeRoleActivity, 
+    assumeRoleInput)
+
+// 3. Extract token from result
+var tokenOutput *routing.AssumeRoleOutput
+output.Get(&tokenOutput)
+
+// 4. Use token in LLM Router
+auth := &routing.LLMAuth{
+    Type:  routing.AuthTypeBearer,
+    Token: tokenOutput.Token,  // ← JWT valid for 30 minutes
 }
+router := routing.NewLLMRouter(config)
 ```
 
-The LLM API server should validate scopes before executing requests.
+### Workflow Pattern: AssumeRole → LLM Router → Activities
 
-### Audit & Compliance
+```go
+// Step 1: Get temporary credentials
+assumeRoleResult := workflow.ExecuteActivity(ctx, routing.AssumeRoleActivity, &routing.AssumeRoleInput{
+    Identity: workflowInput.UserID,
+    Scope:    "llm:read llm:write",
+    DurationSeconds: 1800,
+})
+var token *routing.AssumeRoleOutput
+assumeRoleResult.Get(&token)
 
-All requests include identifying headers for audit trails:
+// Step 2: Use token for all LLM router calls
+routerInput := &routing.LLMRouterInput{
+    Message:  "Analyze code for security",
+    Context:  map[string]interface{}{"repo": "myrepo"},
+}
 
+routerOutput := workflow.ExecuteActivity(ctx, routing.LLMRouterActivity, routerInput)
+// LLMRouter automatically uses the token from LLMRouterConfig
+
+// Step 3: Execute generated workflow with same token
+// (token baked into all activity calls)
 ```
-POST /v1/chat/completions
-Authorization: Bearer eyJhbGc...
-X-Tenant-ID: customer-A
-X-OAuth-Scopes: llm:read
 
-# Server logs:
-# timestamp=2024-01-15T10:30:00Z tenant=customer-A scope=llm:read status=200 tokens=1500
+### Configuration: Credentials from Vault
+
+Never hardcode credentials. Use Kubernetes Secrets or Hashicorp Vault:
+
+```bash
+# In K8s secret
+kubectl create secret generic llm-oauth-creds \
+  --from-literal=OAUTH_CLIENT_ID="client-xxx" \
+  --from-literal=OAUTH_CLIENT_SECRET="secret-yyy" \
+  --from-literal=AUTH_SERVER_URL="https://auth.company.com"
+
+# Pod reads from secret
+env:
+  - name: OAUTH_CLIENT_ID
+    valueFrom:
+      secretKeyRef:
+        name: llm-oauth-creds
+        key: OAUTH_CLIENT_ID
 ```
 
+In code, AssumeRoleActivity reads from environment:
+```go
+input := &routing.AssumeRoleInput{
+    Identity: "[email protected]",
+    Scope:    "llm:read",
+    // clientId, clientSecret, authServerUrl read from env automatically
+}
+result, _ := AssumeRoleActivity(ctx, input)
+```
+
+### Token Lifecycle
+
+| Stage | Duration | Action |
+|-------|----------|--------|
+| **Request** | T+0s | User calls AssumeRoleActivity with identity + scope |
+| **Grant** | T+1s | Auth server validates, issues JWT (default: 1hr validity) |
+| **Use** | T+1s to T+3600s | LLMRouter uses token for all api.riotpiao.com calls |
+| **Refresh** | Before expiry | If workflow > 1hr, request new token via AssumeRole again |
+| **Revoke** | On demand | Auth server can immediately revoke token if needed |
+
+### Scopes & Access Control
+
+Scopes define granular permissions:
+
+```go
+// Read-only access (safe for analytics)
+asScope: "llm:read"
+
+// Full access (for agent workflows)
+scope: "llm:read llm:write"
+
+// Admin access (for operator/setup)
+scope: "llm:admin"
+```
+
+The LLM API validates scopes on every request. AssumeRoleActivity can't escalate privileges—scopes returned by auth server are trusted.
+
 ---
 
 ## Documentation

action/implementer.go

@@ -1,93 +0,0 @@
-package action
-
-import (
-	"context"
-	"fmt"
-
-	"github.com/rockliang/poimen/workflows/action/llm"
-	"github.com/rockliang/poimen/workflows/prompts"
-	"github.com/rockliang/poimen/workflows/statemachine"
-	"go.temporal.io/sdk/activity"
-)
-
-// ImplementerInput is input to ImplementerActivity.
-type ImplementerInput struct {
-	Config      statemachine.OrchestratorConfig
-	TaskID      string
-	WorktreePath string
-	Lessons     string // "known errors — do not repeat" section
-}
-
-// ImplementerOutput is the output of ImplementerActivity.
-type ImplementerOutput struct {
-	Success bool
-	Changes string // summary of changes made
-}
-
-// ImplementerActivity calls the Implementer LLM to implement the task.
-func ImplementerActivity(ctx context.Context, in ImplementerInput) (ImplementerOutput, error) {
-	// Record heartbeat
-	activity.RecordHeartbeat(ctx, "starting implementer for "+in.TaskID)
-
-	// Get LLM client
-	client, err := llm.NewClient()
-	if err != nil {
-		return ImplementerOutput{}, fmt.Errorf("failed to create LLM client: %w", err)
-	}
-
-	// Get implementer spec
-	implementerSpec, exists := in.Config.RolePrompts["implementer"]
-	if !exists {
-		return ImplementerOutput{}, fmt.Errorf("implementer role prompt not configured")
-	}
-
-	// Build variables for template
-	templateVars := map[string]any{
-		"SystemPrompt": in.Config.SystemPrompt,
-		"Task":         in.TaskID,
-		"WorktreePath": in.WorktreePath,
-	}
-
-	// Inject lessons if provided
-	if in.Lessons != "" {
-		templateVars["Lessons"] = in.Lessons
-	}
-
-	// Render template
-	var templateContent string
-	if implementerSpec.RawTemplate != "" {
-		templateContent = implementerSpec.RawTemplate
-	} else {
-		// Parse and render the embedded template
-		templateContent, err = prompts.Render(implementerSpec.TemplateRef, templateVars)
-		if err != nil {
-			return ImplementerOutput{}, fmt.Errorf("failed to render implementer template: %w", err)
-		}
-	}
-
-	// Call LLM
-	messages := []llm.MessageParam{
-		{
-			Role:    "user",
-			Content: templateContent,
-		},
-	}
-
-	response, err := client.CreateMessage(ctx, llm.MessageInput{
-		Model:        implementerSpec.Model,
-		SystemPrompt: in.Config.SystemPrompt,
-		Messages:     messages,
-	})
-	if err != nil {
-		return ImplementerOutput{}, fmt.Errorf("implementer LLM call failed: %w", err)
-	}
-
-	// Record progress
-	activity.RecordHeartbeat(ctx, "implementer completed for "+in.TaskID)
-
-	// Return success (in full implementation would parse response and execute tool calls)
-	return ImplementerOutput{
-		Success: true,
-		Changes: response,
-	}, nil
-}

action/judge.go

@@ -1,78 +0,0 @@
-package action
-
-import (
-	"context"
-	"fmt"
-
-	"github.com/rockliang/poimen/workflows/action/llm"
-	"github.com/rockliang/poimen/workflows/prompts"
-	"github.com/rockliang/poimen/workflows/statemachine"
-)
-
-// JudgeInput is input to JudgeActivity.
-type JudgeInput struct {
-	Config              statemachine.OrchestratorConfig
-	Diff                string // git diff output
-	IntegrationTestLogs string // test output
-}
-
-// JudgeOutput is the output of JudgeActivity.
-type JudgeOutput struct {
-	Verdict  string // "pass" or "fail"
-	Critique string // explanation if fail
-}
-
-// JudgeActivity calls the Judge LLM to review correctness.
-func JudgeActivity(ctx context.Context, in JudgeInput) (JudgeOutput, error) {
-	// Get LLM client
-	client, err := llm.NewClient()
-	if err != nil {
-		return JudgeOutput{}, fmt.Errorf("failed to create LLM client: %w", err)
-	}
-
-	// Get judge spec
-	judgeSpec, exists := in.Config.RolePrompts["judge"]
-	if !exists {
-		return JudgeOutput{}, fmt.Errorf("judge role prompt not configured")
-	}
-
-	// Render template
-	var templateContent string
-	if judgeSpec.RawTemplate != "" {
-		templateContent = judgeSpec.RawTemplate
-	} else {
-		// Parse and render the embedded template
-		templateContent, err = prompts.Render(judgeSpec.TemplateRef, map[string]any{
-			"SystemPrompt": in.Config.SystemPrompt,
-			"Diff":         in.Diff,
-			"TestResult":   in.IntegrationTestLogs,
-		})
-		if err != nil {
-			return JudgeOutput{}, fmt.Errorf("failed to render judge template: %w", err)
-		}
-	}
-
-	// Call LLM
-	messages := []llm.MessageParam{
-		{
-			Role:    "user",
-			Content: templateContent,
-		},
-	}
-
-	response, err := client.CreateMessage(ctx, llm.MessageInput{
-		Model:        judgeSpec.Model,
-		SystemPrompt: in.Config.SystemPrompt,
-		Messages:     messages,
-	})
-	if err != nil {
-		return JudgeOutput{}, fmt.Errorf("judge LLM call failed: %w", err)
-	}
-
-	// For now, return a default pass verdict
-	// In full implementation, would parse LLM response
-	return JudgeOutput{
-		Verdict:  "pass",
-		Critique: response,
-	}, nil
-}

action/planner.go

@@ -1,91 +0,0 @@
-package action
-
-import (
-	"context"
-	"fmt"
-
-	"github.com/rockliang/poimen/workflows/action/llm"
-	"github.com/rockliang/poimen/workflows/prompts"
-	"github.com/rockliang/poimen/workflows/statemachine"
-)
-
-// PlanningInput is input to PlanningActivity.
-type PlanningInput struct {
-	Config        statemachine.OrchestratorConfig
-	BoardState    string                            // JSON or markdown of task board
-	RepoPath      string                            // Path to target repository
-	Milestone     string                            // e.g., "T0"
-	TaskResults   []statemachine.TaskUnitOutput     // Results from completed tasks
-}
-
-// TaskDispatch represents a dispatched task.
-type TaskDispatch struct {
-	TaskID       string
-	PromptSpec   statemachine.PromptSpec
-	BaseTimeout  *int64 // optional override in milliseconds
-}
-
-// PlanningOutput is the output of PlanningActivity.
-type PlanningOutput struct {
-	TasksToDispatch        []string // Task IDs to dispatch in this cycle
-	CompletedBranches      []string // Branches to squash merge (when milestone complete)
-	SubmilestoneComplete   bool     // Whether the milestone is complete
-}
-
-// PlanningActivity calls the Planner LLM to decide which tasks to dispatch.
-func PlanningActivity(ctx context.Context, in PlanningInput) (PlanningOutput, error) {
-	// Get LLM client
-	client, err := llm.NewClient()
-	if err != nil {
-		return PlanningOutput{}, fmt.Errorf("failed to create LLM client: %w", err)
-	}
-
-	// Get planner spec
-	plannerSpec, exists := in.Config.RolePrompts["planner"]
-	if !exists {
-		return PlanningOutput{}, fmt.Errorf("planner role prompt not configured")
-	}
-
-	// Render template
-	var templateContent string
-	if plannerSpec.RawTemplate != "" {
-		templateContent = plannerSpec.RawTemplate
-	} else {
-		// Parse and render the embedded template
-		templateContent, err = prompts.Render(plannerSpec.TemplateRef, map[string]any{
-			"SystemPrompt": in.Config.SystemPrompt,
-			"BoardState":   in.BoardState,
-			"Milestone":    in.Milestone,
-			"Config":       in.Config,
-		})
-		if err != nil {
-			return PlanningOutput{}, fmt.Errorf("failed to render planner template: %w", err)
-		}
-	}
-
-	// Call LLM
-	messages := []llm.MessageParam{
-		{
-			Role:    "user",
-			Content: templateContent,
-		},
-	}
-
-	response, err := client.CreateMessage(ctx, llm.MessageInput{
-		Model:        plannerSpec.Model,
-		SystemPrompt: in.Config.SystemPrompt,
-		Messages:     messages,
-	})
-	if err != nil {
-		return PlanningOutput{}, fmt.Errorf("planner LLM call failed: %w", err)
-	}
-
-	// For now, return empty dispatch (will be parsed from LLM response in full implementation)
-	// This is a stub that allows the test to verify the activity is called
-	_ = response
-	return PlanningOutput{
-		TasksToDispatch:      []string{},
-		CompletedBranches:    []string{},
-		SubmilestoneComplete: false,
-	}, nil
-}

activity/analysis.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/analysis_test.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/assume_role.go

@@ -0,0 +1,216 @@
+package activity
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"time"
+)
+
+// AssumeRoleInput is the input to AssumeRoleActivity
+type AssumeRoleInput struct {
+	// Identity is the user/service identity requesting access
+	Identity string `json:"identity"`
+	
+	// ClientID is the OAuth2/OIDC client ID (from vault or env)
+	ClientID string `json:"clientId,omitempty"`
+	
+	// ClientSecret is the OAuth2/OIDC client secret (from vault or env)
+	ClientSecret string `json:"clientSecret,omitempty"`
+	
+	// Scope defines what APIs this token can access (e.g., "llm:read llm:write")
+	Scope string `json:"scope"`
+	
+	// DurationSeconds is how long the token is valid (default: 3600 = 1 hour)
+	DurationSeconds int `json:"durationSeconds,omitempty"`
+	
+	// AuthServerURL is the auth server endpoint (from env if not provided)
+	AuthServerURL string `json:"authServerUrl,omitempty"`
+}
+
+// AssumeRoleOutput is the output from AssumeRoleActivity
+type AssumeRoleOutput struct {
+	// Token is the JWT token for calling api.riotpiao.com
+	Token string `json:"token"`
+	
+	// ExpiresAt is when the token expires (Unix timestamp)
+	ExpiresAt int64 `json:"expiresAt"`
+	
+	// ExpiresIn is the duration in seconds until expiration
+	ExpiresIn int `json:"expiresIn"`
+	
+	// TokenType is typically "Bearer"
+	TokenType string `json:"tokenType"`
+}
+
+// oauthTokenRequest is sent to the auth server
+type oauthTokenRequest struct {
+	GrantType    string `json:"grant_type"`
+	ClientID     string `json:"client_id"`
+	ClientSecret string `json:"client_secret"`
+	Scope        string `json:"scope"`
+	Subject      string `json:"subject,omitempty"` // The identity being assumed
+}
+
+// oauthTokenResponse is returned from the auth server
+type oauthTokenResponse struct {
+	AccessToken string `json:"access_token"`
+	TokenType   string `json:"token_type"`
+	ExpiresIn   int    `json:"expires_in"`
+	Scope       string `json:"scope"`
+}
+
+// AssumeRoleActivity requests a temporary JWT token for accessing LLM APIs
+//
+// This activity works like AWS AssumeRole:
+// 1. User provides identity + scope of access needed
+// 2. Activity exchanges credentials with auth server
+// 3. Returns JWT token valid for a limited time
+// 4. Caller uses token in subsequent LLM API calls
+//
+// Security: Credentials should come from vault/secrets, never hardcoded
+func AssumeRoleActivity(ctx context.Context, input *AssumeRoleInput) (*AssumeRoleOutput, error) {
+	// Validate inputs
+	if err := validateAssumeRoleInput(input); err != nil {
+		return nil, err
+	}
+
+	// Resolve configuration from input + environment
+	config, err := resolveAssumeRoleConfig(input)
+	if err != nil {
+		return nil, err
+	}
+
+	// Request token from auth server
+	tokenResp, err := requestAuthToken(ctx, config, input)
+	if err != nil {
+		return nil, err
+	}
+
+	// Build output
+	return buildAssumeRoleOutput(tokenResp), nil
+}
+
+// validateAssumeRoleInput checks required fields
+func validateAssumeRoleInput(input *AssumeRoleInput) error {
+	if input.Identity == "" {
+		return fmt.Errorf("identity is required")
+	}
+	if input.Scope == "" {
+		return fmt.Errorf("scope is required (e.g., 'llm:read' or 'llm:read llm:write')")
+	}
+	return nil
+}
+
+// assumeRoleConfig holds resolved configuration
+type assumeRoleConfig struct {
+	authServerURL string
+	clientID      string
+	clientSecret  string
+	duration      int
+}
+
+// resolveAssumeRoleConfig gets config from input or environment
+func resolveAssumeRoleConfig(input *AssumeRoleInput) (*assumeRoleConfig, error) {
+	cfg := &assumeRoleConfig{}
+
+	// Helper function to avoid DRY violation
+	getOrEnv := func(val, envKey, fieldName string) (string, error) {
+		if val != "" {
+			return val, nil
+		}
+		if val = os.Getenv(envKey); val != "" {
+			return val, nil
+		}
+		return "", fmt.Errorf("%s not provided and %s not set", fieldName, envKey)
+	}
+
+	var err error
+	if cfg.authServerURL, err = getOrEnv(input.AuthServerURL, "AUTH_SERVER_URL", "authServerUrl"); err != nil {
+		return nil, err
+	}
+	if cfg.clientID, err = getOrEnv(input.ClientID, "OAUTH_CLIENT_ID", "clientId"); err != nil {
+		return nil, err
+	}
+	if cfg.clientSecret, err = getOrEnv(input.ClientSecret, "OAUTH_CLIENT_SECRET", "clientSecret"); err != nil {
+		return nil, err
+	}
+
+	// Validate and set duration
+	cfg.duration = input.DurationSeconds
+	if cfg.duration == 0 {
+		cfg.duration = 3600 // 1 hour default
+	}
+	if cfg.duration > 86400 {
+		cfg.duration = 86400 // Max 24 hours
+	}
+
+	return cfg, nil
+}
+
+// requestAuthToken calls the auth server and returns the token response
+func requestAuthToken(ctx context.Context, config *assumeRoleConfig, input *AssumeRoleInput) (*oauthTokenResponse, error) {
+	tokenReq := oauthTokenRequest{
+		GrantType:    "client_credentials",
+		ClientID:     config.clientID,
+		ClientSecret: config.clientSecret,
+		Scope:        input.Scope,
+		Subject:      input.Identity,
+	}
+
+	reqBody, err := json.Marshal(tokenReq)
+	if err != nil {
+		return nil, fmt.Errorf("failed to marshal token request: %w", err)
+	}
+
+	httpReq, err := http.NewRequestWithContext(ctx, "POST",
+		fmt.Sprintf("%s/oauth/token", config.authServerURL),
+		bytes.NewReader(reqBody))
+	if err != nil {
+		return nil, fmt.Errorf("failed to create request: %w", err)
+	}
+
+	httpReq.Header.Set("Content-Type", "application/json")
+
+	client := &http.Client{Timeout: 10 * time.Second}
+	resp, err := client.Do(httpReq)
+	if err != nil {
+		return nil, fmt.Errorf("failed to call auth server: %w", err)
+	}
+	defer resp.Body.Close()
+
+	respBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read response: %w", err)
+	}
+
+	if resp.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("auth server returned status %d: %s", resp.StatusCode, string(respBody))
+	}
+
+	var tokenResp oauthTokenResponse
+	if err := json.Unmarshal(respBody, &tokenResp); err != nil {
+		return nil, fmt.Errorf("failed to unmarshal token response: %w", err)
+	}
+
+	if tokenResp.AccessToken == "" {
+		return nil, fmt.Errorf("auth server returned empty access token")
+	}
+
+	return &tokenResp, nil
+}
+
+// buildAssumeRoleOutput constructs the output from token response
+func buildAssumeRoleOutput(tokenResp *oauthTokenResponse) *AssumeRoleOutput {
+	expiresAt := time.Now().Add(time.Duration(tokenResp.ExpiresIn) * time.Second).Unix()
+	return &AssumeRoleOutput{
+		Token:     tokenResp.AccessToken,
+		ExpiresAt: expiresAt,
+		ExpiresIn: tokenResp.ExpiresIn,
+		TokenType: tokenResp.TokenType,
+	}
+}

activity/canvas_compatibility.go

@@ -0,0 +1,359 @@
+package activity
+
+import (
+	"encoding/json"
+	"fmt"
+	"strings"
+
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// CanvasCompatibilityOutput validation results
+type CanvasCompatibilityOutput struct {
+	IsValid           bool                    `json:"is_valid"`
+	Incompatibilities []IncompatibilityWarning `json:"incompatibilities"`
+	DisconnectedNodes []string                 `json:"disconnected_nodes"`
+	Warnings          []string                 `json:"warnings"`
+}
+
+// IncompatibilityWarning explains why two activities can't be connected
+type IncompatibilityWarning struct {
+	Source       string `json:"source"`        // Source node ID
+	Target       string `json:"target"`        // Target node ID
+	Reason       string `json:"reason"`        // Why they can't connect
+	SourceNeeds  string `json:"source_needs"`  // What source would need to output
+	TargetNeeds  string `json:"target_needs"`  // What target requires as input
+	Suggestion   string `json:"suggestion"`    // Suggestion to make it work
+}
+
+// ActivitySchema describes what an activity needs/provides
+type ActivitySchema struct {
+	ActivityType string                 `json:"activity_type"`
+	Inputs       map[string]InputField  `json:"inputs"`
+	Outputs      map[string]OutputField `json:"outputs"`
+}
+
+type InputField struct {
+	Type        string   `json:"type"`
+	Description string   `json:"description"`
+	Required    bool     `json:"required"`
+	Enum        []string `json:"enum,omitempty"`
+}
+
+type OutputField struct {
+	Type        string `json:"type"`
+	Description string `json:"description"`
+}
+
+// getActivitySchema returns schema from knowledge base
+func getActivitySchema(activityType string) (*ActivitySchema, error) {
+	kb := knowledgeBaseData()
+	if kb == "" {
+		return nil, fmt.Errorf("knowledge base not loaded")
+	}
+
+	var activities []map[string]interface{}
+	if err := json.Unmarshal([]byte(kb), &activities); err != nil {
+		// Try to extract activities from full KB structure
+		var fullKB map[string]interface{}
+		if err := json.Unmarshal([]byte(kb), &fullKB); err != nil {
+			return nil, fmt.Errorf("failed to parse knowledge base")
+		}
+		if activitiesRaw, ok := fullKB["activities"]; ok {
+			if b, err := json.Marshal(activitiesRaw); err == nil {
+				if err := json.Unmarshal(b, &activities); err != nil {
+					return nil, fmt.Errorf("failed to extract activities from KB")
+				}
+			}
+		}
+	}
+
+	// Find matching activity
+	for _, act := range activities {
+		if name, ok := act["name"].(string); ok {
+			if toActivityName(activityType) == name {
+				// Convert to ActivitySchema
+				schema := &ActivitySchema{
+					ActivityType: activityType,
+					Inputs:       make(map[string]InputField),
+					Outputs:      make(map[string]OutputField),
+				}
+
+				if inputs, ok := act["inputs"].(map[string]interface{}); ok {
+					for key, val := range inputs {
+						if field, ok := val.(map[string]interface{}); ok {
+							schema.Inputs[key] = parseInputField(field)
+						}
+					}
+				}
+
+				if outputs, ok := act["outputs"].(map[string]interface{}); ok {
+					for key, val := range outputs {
+						if field, ok := val.(map[string]interface{}); ok {
+							schema.Outputs[key] = parseOutputField(field)
+						}
+					}
+				}
+
+				return schema, nil
+			}
+		}
+	}
+
+	return nil, fmt.Errorf("activity %s not found in knowledge base", activityType)
+}
+
+func parseInputField(data map[string]interface{}) InputField {
+	field := InputField{}
+	if t, ok := data["type"].(string); ok {
+		field.Type = t
+	}
+	if d, ok := data["description"].(string); ok {
+		field.Description = d
+	}
+	if r, ok := data["required"].(bool); ok {
+		field.Required = r
+	}
+	return field
+}
+
+func parseOutputField(data map[string]interface{}) OutputField {
+	field := OutputField{}
+	if t, ok := data["type"].(string); ok {
+		field.Type = t
+	}
+	if d, ok := data["description"].(string); ok {
+		field.Description = d
+	}
+	return field
+}
+
+// CheckConnectionCompatibility validates if source can connect to target
+func CheckConnectionCompatibility(sourceNode, targetNode db.WorkflowNode) []IncompatibilityWarning {
+	warnings := []IncompatibilityWarning{}
+
+	sourceSchema, err := getActivitySchema(sourceNode.Type)
+	if err != nil {
+		warnings = append(warnings, IncompatibilityWarning{
+			Source:     sourceNode.ID,
+			Target:     targetNode.ID,
+			Reason:     fmt.Sprintf("Source activity schema not found: %v", err),
+			Suggestion: "Ensure source activity type is registered in knowledge base",
+		})
+		return warnings
+	}
+
+	targetSchema, err := getActivitySchema(targetNode.Type)
+	if err != nil {
+		warnings = append(warnings, IncompatibilityWarning{
+			Source:     sourceNode.ID,
+			Target:     targetNode.ID,
+			Reason:     fmt.Sprintf("Target activity schema not found: %v", err),
+			Suggestion: "Ensure target activity type is registered in knowledge base",
+		})
+		return warnings
+	}
+
+	// Check if source produces outputs that target can consume
+	if len(sourceSchema.Outputs) == 0 {
+		warnings = append(warnings, IncompatibilityWarning{
+			Source:     sourceNode.ID,
+			Target:     targetNode.ID,
+			Reason:     fmt.Sprintf("%s produces no outputs", sourceNode.Type),
+			SourceNeeds: "any output",
+			Suggestion: "Source activity must produce outputs",
+		})
+		return warnings
+	}
+
+	if len(targetSchema.Inputs) == 0 {
+		warnings = append(warnings, IncompatibilityWarning{
+			Source:     sourceNode.ID,
+			Target:     targetNode.ID,
+			Reason:     fmt.Sprintf("%s accepts no inputs", targetNode.Type),
+			TargetNeeds: "no input",
+			Suggestion: "Target activity must accept inputs. Check if it's a terminal activity.",
+		})
+		return warnings
+	}
+
+	// Match outputs to inputs
+	sourceOutputs := getOutputNames(sourceSchema.Outputs)
+	targetInputs := getInputNames(targetSchema.Inputs)
+
+	if len(sourceOutputs) == 0 || len(targetInputs) == 0 {
+		warnings = append(warnings, IncompatibilityWarning{
+			Source:       sourceNode.ID,
+			Target:       targetNode.ID,
+			Reason:       "No compatible output/input fields found",
+			SourceNeeds:  strings.Join(sourceOutputs, ", "),
+			TargetNeeds:  strings.Join(targetInputs, ", "),
+			Suggestion:   "Use LLM transformation to map outputs to inputs",
+		})
+	}
+
+	return warnings
+}
+
+// CheckCanvasConnectivity analyzes all suggested edges for compatibility
+func CheckCanvasConnectivity(nodes []db.WorkflowNode, suggestedEdges []EdgeWithWording) []IncompatibilityWarning {
+	warnings := []IncompatibilityWarning{}
+	nodeMap := make(map[string]db.WorkflowNode)
+	for _, n := range nodes {
+		nodeMap[n.ID] = n
+	}
+
+	for _, edge := range suggestedEdges {
+		sourceNode, ok := nodeMap[edge.Source]
+		if !ok {
+			continue
+		}
+		targetNode, ok := nodeMap[edge.Target]
+		if !ok {
+			continue
+		}
+
+		edgeWarnings := CheckConnectionCompatibility(sourceNode, targetNode)
+		warnings = append(warnings, edgeWarnings...)
+	}
+
+	return warnings
+}
+
+// IdentifyDisconnectedNodes finds nodes that can't connect to anything
+func IdentifyDisconnectedNodes(nodes []db.WorkflowNode, suggestedEdges []EdgeWithWording) []string {
+	edgeMap := make(map[string]bool)
+	for _, edge := range suggestedEdges {
+		edgeMap[edge.Source] = true
+		edgeMap[edge.Target] = true
+	}
+
+	var disconnected []string
+	for _, node := range nodes {
+		if !edgeMap[node.ID] {
+			disconnected = append(disconnected, node.ID)
+		}
+	}
+	return disconnected
+}
+
+// toActivityName converts canvas type to activity name (e.g., "clone-repo" -> "CloneRepoActivity")
+func toActivityName(canvasType string) string {
+	parts := strings.Split(canvasType, "-")
+	var result string
+	for _, part := range parts {
+		if part != "" {
+			result += strings.ToUpper(part[:1]) + strings.ToLower(part[1:])
+		}
+	}
+	return result + "Activity"
+}
+
+// getOutputNames extracts output field names
+func getOutputNames(outputs map[string]OutputField) []string {
+	var names []string
+	for name := range outputs {
+		names = append(names, name)
+	}
+	return names
+}
+
+// getInputNames extracts input field names (required ones highlighted)
+func getInputNames(inputs map[string]InputField) []string {
+	var names []string
+	for name, field := range inputs {
+		if field.Required {
+			names = append(names, name+"*")
+		} else {
+			names = append(names, name)
+		}
+	}
+	return names
+}
+
+// SuggestDataTransformation proposes how to connect incompatible activities
+func SuggestDataTransformation(sourceNode, targetNode db.WorkflowNode) string {
+	sourceSchema, _ := getActivitySchema(sourceNode.Type)
+	targetSchema, _ := getActivitySchema(targetNode.Type)
+
+	if sourceSchema == nil || targetSchema == nil {
+		return "Cannot analyze compatibility without schemas"
+	}
+
+	sourceOuts := getOutputNames(sourceSchema.Outputs)
+	targetIns := getInputNames(targetSchema.Inputs)
+
+	return fmt.Sprintf(
+		"To connect %s → %s:\n"+
+			"  %s outputs: %s\n"+
+			"  %s needs: %s\n"+
+			"  Solution: Use LLM transformation node to map outputs to inputs",
+		sourceNode.Label, targetNode.Label,
+		sourceNode.Type, strings.Join(sourceOuts, ", "),
+		targetNode.Type, strings.Join(targetIns, ", "),
+	)
+}
+
+// knowledgeBaseData returns raw KB JSON (stub - implement with actual KB loading)
+func knowledgeBaseData() string {
+	// This would load from activity_knowledge_base.json
+	// For now, return empty - real implementation loads from file
+	return ""
+}
+
+// CanvasCompatibilityActivity validates workflow canvas for type mismatches and isolation
+func CanvasCompatibilityActivity(ctx interface{}, input CanvasCompatibilityInput) (CanvasCompatibilityOutput, error) {
+	output := CanvasCompatibilityOutput{
+		IsValid:           true,
+		Incompatibilities: []IncompatibilityWarning{},
+		DisconnectedNodes: []string{},
+		Warnings:          []string{},
+	}
+
+	// Check all edges for compatibility
+	for _, edge := range input.Edges {
+		var sourceNode, targetNode *db.WorkflowNode
+		for i := range input.Nodes {
+			if input.Nodes[i].ID == edge.Source {
+				sourceNode = &input.Nodes[i]
+			}
+			if input.Nodes[i].ID == edge.Target {
+				targetNode = &input.Nodes[i]
+			}
+		}
+
+		if sourceNode != nil && targetNode != nil {
+			if warning, err := ValidateConnection(sourceNode, targetNode); err != nil {
+				output.IsValid = false
+				output.Incompatibilities = append(output.Incompatibilities, warning)
+			}
+		}
+	}
+
+	// Find disconnected nodes
+	connected := make(map[string]bool)
+	for _, edge := range input.Edges {
+		connected[edge.Source] = true
+		connected[edge.Target] = true
+	}
+
+	for _, node := range input.Nodes {
+		if node.Type == "activity" && !connected[node.ID] {
+			output.DisconnectedNodes = append(output.DisconnectedNodes, node.ID)
+		}
+	}
+
+	return output, nil
+}
+
+// ValidateConnection checks if two nodes can be connected based on their types.
+func ValidateConnection(source, target *db.WorkflowNode) (IncompatibilityWarning, error) {
+	if source.Type != "activity" || target.Type != "activity" {
+		return IncompatibilityWarning{
+			Source: source.ID,
+			Target: target.ID,
+			Reason: fmt.Sprintf("Cannot connect %s to %s: both must be activity type", source.Type, target.Type),
+		}, fmt.Errorf("type mismatch")
+	}
+	return IncompatibilityWarning{}, nil
+}

activity/canvas_reasoner.go

@@ -0,0 +1,253 @@
+package activity
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+
+	"github.com/rockliang/poimen/workflows/activity/llm"
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// CanvasReasonerOutput returns suggested edges and reasoning
+type CanvasReasonerOutput struct {
+	SuggestedEdges        []EdgeWithWording             `json:"suggested_edges"`         // Edges with wording
+	RemovedEdges          []db.WorkflowEdge             `json:"removed_edges,omitempty"` // Edges to remove
+	Reasoning             string                        `json:"reasoning"`               // LLM explanation
+	Confidence            float64                       `json:"confidence"`              // 0.0-1.0
+	IncompatibleEdges     []IncompatibilityWarning      `json:"incompatible_edges,omitempty"` // Can't connect
+	DisconnectedNodes     []string                      `json:"disconnected_nodes,omitempty"` // No connections
+	UserAlerts            []string                      `json:"user_alerts,omitempty"`        // Human-readable warnings
+}
+
+// CanvasReasonerActivity uses LLM to infer connections between workflow activities
+func CanvasReasonerActivity(ctx context.Context, in CanvasReasonerInput) (CanvasReasonerOutput, error) {
+	logger := newActivityLogger(ctx)
+
+	output := CanvasReasonerOutput{
+		SuggestedEdges: []EdgeWithWording{},
+	}
+
+	if len(in.Nodes) == 0 {
+		return output, fmt.Errorf("no nodes provided")
+	}
+
+	logger.Info("Analyzing canvas with %d nodes, %d edges", len(in.Nodes), len(in.Edges))
+
+	// Build activity descriptions for LLM context
+	nodeDesc := buildNodeDescriptions(in.Nodes)
+	edgeDesc := buildEdgeDescriptions(in.Edges)
+
+	// Create prompt for LLM reasoning with relation wording
+	systemPrompt := `You are a workflow automation expert. Analyze activities and suggest logical connections with semantic descriptions.
+
+CRITICAL RULES:
+1. Only suggest edges where outputs→inputs match
+2. Provide relation wording: verb, source_output, target_input
+3. Assess connection confidence (0.0-1.0)
+4. Flag type mismatches that need transformers
+
+Respond with JSON:
+{
+  "edges": [
+    {
+      "source": "node-1",
+      "target": "node-2",
+      "relation_type": "data-flow|dependency|conditional|parallel",
+      "relation_label": "Node1 outputs X → Node2 requires X",
+      "relation_wording": {
+        "verb": "outputs|depends-on|triggers|etc",
+        "source_output": "field_name (type): description",
+        "target_input": "field_name (type, required?): description",
+        "connection_type": "direct-map|requires-transformer|conditional",
+        "confidence": 0.95,
+        "semantic_match": "Explanation of why this makes sense"
+      }
+    }
+  ],
+  "reasoning": "Overall workflow structure explanation",
+  "confidence": 0.85
+}`
+
+	userPrompt := fmt.Sprintf(`Canvas Analysis:
+
+Nodes (including inputs/outputs):
+%s
+
+Current Edges:
+%s
+
+Task: %s
+
+KEY RULES:
+- Preserve existing edges and suggest only NEW edges to add
+- SKIP any connections where input/output types don't match
+- If an activity has no outputs, it cannot be a source
+- If an activity has no inputs, it cannot be a target
+- Note any activities that are hard to connect (terminal activities, generators, etc)
+
+Return ONLY valid JSON, no markdown code blocks.`, nodeDesc, edgeDesc, getReasoningTask(in.PreserveExisting))
+
+	logger.Info("Calling LLM reasoning (preserve_existing=%v)", in.PreserveExisting)
+
+	// Call LLM
+	client, err := llm.NewClient()
+	if err != nil {
+		return output, fmt.Errorf("failed to create LLM client: %w", err)
+	}
+
+	response, err := client.CreateMessage(ctx, llm.MessageInput{
+		Model: ModelSpec{
+			ModelID: "reasoning", // Use reasoning model for complex analysis
+		},
+		SystemPrompt: systemPrompt,
+		Messages: []llm.MessageParam{
+			{
+				Role:    "user",
+				Content: userPrompt,
+			},
+		},
+		AuthToken: in.AuthToken,
+	})
+
+	if err != nil {
+		return output, fmt.Errorf("LLM reasoning failed: %w", err)
+	}
+
+	// Parse LLM response
+	var reasonerResp struct {
+		Edges      []EdgeWithWording `json:"edges"`
+		Reasoning  string            `json:"reasoning"`
+		Confidence float64           `json:"confidence"`
+	}
+
+	if err := json.Unmarshal([]byte(response), &reasonerResp); err != nil {
+		logger.Warn("Failed to parse LLM response as JSON: %v", err)
+		// Try to extract from response text
+		output.Reasoning = response
+		output.Confidence = 0.5
+		return output, fmt.Errorf("failed to parse LLM response: %w", err)
+	}
+
+	// Validate suggested edges
+	nodeMap := make(map[string]bool)
+	for _, n := range in.Nodes {
+		nodeMap[n.ID] = true
+	}
+
+	validEdges := []EdgeWithWording{}
+	for _, edge := range reasonerResp.Edges {
+		if !nodeMap[edge.Source] {
+			logger.Warn("Suggested edge references unknown source: %s", edge.Source)
+			continue
+		}
+		if !nodeMap[edge.Target] {
+			logger.Warn("Suggested edge references unknown target: %s", edge.Target)
+			continue
+		}
+		// Don't suggest self-loops
+		if edge.Source == edge.Target {
+			logger.Warn("Skipping self-loop: %s", edge.Source)
+			continue
+		}
+		validEdges = append(validEdges, edge)
+	}
+
+	output.SuggestedEdges = validEdges
+	output.Reasoning = reasonerResp.Reasoning
+	output.Confidence = reasonerResp.Confidence
+
+	// Check compatibility of suggested edges
+	incompatibilities := CheckCanvasConnectivity(in.Nodes, validEdges)
+	if len(incompatibilities) > 0 {
+		output.IncompatibleEdges = incompatibilities
+		logger.Warn("Found %d incompatible edge connections", len(incompatibilities))
+
+		// Generate user-friendly alerts
+		for i, incompat := range incompatibilities {
+			if i < 5 { // Limit to 5 alerts to avoid spam
+				alert := fmt.Sprintf(
+					"⚠️  %s → %s: %s. %s",
+					incompat.Source, incompat.Target, incompat.Reason, incompat.Suggestion,
+				)
+				output.UserAlerts = append(output.UserAlerts, alert)
+			}
+		}
+	}
+
+	// Identify disconnected nodes
+	disconnected := IdentifyDisconnectedNodes(in.Nodes, validEdges)
+	if len(disconnected) > 0 {
+		output.DisconnectedNodes = disconnected
+		logger.Warn("Found %d disconnected nodes", len(disconnected))
+
+		for _, nodeID := range disconnected {
+			var label string
+			for _, node := range in.Nodes {
+				if node.ID == nodeID {
+					label = node.Label
+					break
+				}
+			}
+			alert := fmt.Sprintf(
+				"🔌 Node '%s' has no connections. Consider adding edges or removing it.",
+				label,
+			)
+			output.UserAlerts = append(output.UserAlerts, alert)
+		}
+	}
+
+	logger.Info("LLM suggested %d edges with confidence %.2f | %d incompatibilities | %d disconnected",
+		len(validEdges), output.Confidence, len(incompatibilities), len(disconnected))
+
+	return output, nil
+}
+
+// buildNodeDescriptions creates readable node descriptions for LLM (including schemas)
+func buildNodeDescriptions(nodes []db.WorkflowNode) string {
+	var desc string
+	for i, node := range nodes {
+		desc += fmt.Sprintf("%d. [%s] %s (type: %s)\n", i+1, node.ID, node.Label, node.Type)
+
+		// Add input/output schema info
+		if schema, err := getActivitySchema(node.Type); err == nil {
+			if len(schema.Inputs) > 0 {
+				desc += fmt.Sprintf("   INPUTS: %v\n", getInputNames(schema.Inputs))
+			} else {
+				desc += fmt.Sprintf("   INPUTS: none (generator/trigger)\n")
+			}
+			if len(schema.Outputs) > 0 {
+				desc += fmt.Sprintf("   OUTPUTS: %v\n", getOutputNames(schema.Outputs))
+			} else {
+				desc += fmt.Sprintf("   OUTPUTS: none (terminal/sink)\n")
+			}
+		}
+
+		if node.Data != nil {
+			if b, err := json.MarshalIndent(node.Data, "   ", "  "); err == nil {
+				desc += fmt.Sprintf("   CONFIG: %s\n", string(b))
+			}
+		}
+	}
+	return desc
+}
+
+// buildEdgeDescriptions creates readable edge descriptions for LLM
+func buildEdgeDescriptions(edges []db.WorkflowEdge) string {
+	if len(edges) == 0 {
+		return "None"
+	}
+	var desc string
+	for i, edge := range edges {
+		desc += fmt.Sprintf("%d. %s → %s\n", i+1, edge.Source, edge.Target)
+	}
+	return desc
+}
+
+// getReasoningTask returns task description based on preservation mode
+func getReasoningTask(preserveExisting bool) string {
+	if preserveExisting {
+		return "Keep all existing edges and suggest ONLY NEW edges to improve workflow"
+	}
+	return "Design optimal workflow by suggesting all connections and noting any redundant edges"
+}

activity/fetch_canvas_relations.go

@@ -0,0 +1,80 @@
+package activity
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// FetchCanvasRelationsActivity fetches canvas + relations from DB
+func FetchCanvasRelationsActivity(ctx context.Context, input FetchCanvasRelationsInput) (CanvasWithRelationsData, error) {
+	logger := newActivityLogger(ctx)
+	output := CanvasWithRelationsData{
+		WorkflowID: input.WorkflowID,
+		Version:    input.Version,
+		Nodes:      []db.WorkflowNode{},
+		Edges:      []db.WorkflowEdge{},
+		Relations:  []EdgeWithWording{},
+	}
+
+	logger.Info("Fetching canvas relations: %s v%d", input.WorkflowID, input.Version)
+
+	// Get database client from context or activity manager
+	dbClient, ok := ctx.Value("db_client").(*db.DB)
+	if !ok {
+		return output, fmt.Errorf("database client not in context")
+	}
+
+	// Fetch workflow
+	workflow, err := dbClient.FetchWorkflow(ctx, input.WorkflowID, "")
+	if err != nil {
+		return output, fmt.Errorf("failed to get workflow: %w", err)
+	}
+
+	// Parse canvas nodes and edges
+	var nodes []db.WorkflowNode
+	if err := json.Unmarshal([]byte(workflow.Nodes), &nodes); err != nil {
+		return output, fmt.Errorf("failed to parse nodes: %w", err)
+	}
+
+	var edges []db.WorkflowEdge
+	if err := json.Unmarshal([]byte(workflow.Edges), &edges); err != nil {
+		return output, fmt.Errorf("failed to parse edges: %w", err)
+	}
+
+	output.Nodes = nodes
+	output.Edges = edges
+	output.UpdatedAt = workflow.UpdatedAt.String()
+
+	// Fetch workflow relations
+	relations, err := dbClient.GetWorkflowRelations(ctx, input.WorkflowID, input.Version)
+	if err != nil {
+		// Relations may not exist for old canvases - this is OK
+		logger.Warn("Failed to fetch relations: %v", err)
+		return output, nil
+	}
+
+	// Map to EdgeWithWording
+	for _, rel := range relations {
+		edge := EdgeWithWording{
+			ID:             rel.ID,
+			Source:         rel.SourceNodeID,
+			Target:         rel.TargetNodeID,
+			RelationType:   rel.RelationType,
+			RelationLabel:  rel.Label,
+			CreatedAt:      rel.CreatedAt.String(),
+		}
+
+		// Parse relation wording JSON
+		if err := json.Unmarshal(rel.RelationWording, &edge.RelationWording); err != nil {
+			logger.Warn("Failed to parse relation wording: %v", err)
+		}
+
+		output.Relations = append(output.Relations, edge)
+	}
+
+	logger.Info("Fetched %d nodes, %d edges, %d relations", len(output.Nodes), len(output.Edges), len(output.Relations))
+	return output, nil
+}

activity/git.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/implementer.go

@@ -0,0 +1,41 @@
+package activity
+
+import (
+	"context"
+
+	"github.com/rockliang/poimen/workflows/pkg/types"
+	"go.temporal.io/sdk/activity"
+)
+
+type ImplementerInput struct {
+	Config       types.OrchestratorConfig
+	TaskID       string
+	WorktreePath string
+	Lessons      string
+}
+
+type ImplementerOutput struct {
+	Success bool
+	Changes string
+}
+
+func ImplementerActivity(ctx context.Context, in ImplementerInput) (ImplementerOutput, error) {
+	activity.RecordHeartbeat(ctx, "starting implementer for "+in.TaskID)
+
+	vars := map[string]any{
+		"SystemPrompt": in.Config.SystemPrompt,
+		"Task":         in.TaskID,
+		"WorktreePath": in.WorktreePath,
+	}
+	if in.Lessons != "" {
+		vars["Lessons"] = in.Lessons
+	}
+
+	response, err := CallRoleLLM(ctx, in.Config, "implementer", vars)
+	if err != nil {
+		return ImplementerOutput{}, err
+	}
+
+	activity.RecordHeartbeat(ctx, "implementer completed for "+in.TaskID)
+	return ImplementerOutput{Success: true, Changes: response}, nil
+}

activity/index_graph_rag.go

@@ -0,0 +1,51 @@
+package activity
+
+import (
+	"context"
+	"time"
+)
+
+// IndexGraphRAGActivity indexes workflow canvas to GraphRAG (stub for now)
+func IndexGraphRAGActivity(ctx context.Context, input IndexGraphRAGInput) (IndexGraphRAGOutput, error) {
+	output := IndexGraphRAGOutput{
+		WorkflowID:      input.WorkflowID,
+		Version:         input.Version,
+		Status:          "indexed",
+		IndexedEntities: len(input.Nodes),
+		IndexedEdges:    len(input.Relations),
+		IndexedAt:       time.Now().UTC().Format(time.RFC3339),
+	}
+
+	// Stub implementation - actual GraphRAG indexing would happen here
+	// For now, just return success
+	return output, nil
+}
+
+// QueryGraphRAGRelationsInput for direct relation discovery
+type QueryGraphRAGRelationsInput struct {
+	WorkflowID string                 `json:"workflow_id"`
+	Version    int                    `json:"version"`
+	Query      string                 `json:"query"`
+	TopK       int                    `json:"top_k"`
+	Filters    map[string]interface{} `json:"filters,omitempty"`
+}
+
+// QueryGraphRAGRelationsOutput returns discovered relations
+type QueryGraphRAGRelationsOutput struct {
+	Query       string            `json:"query"`
+	Results     []EdgeWithWording `json:"results"`
+	TotalCount  int               `json:"total_count"`
+	ExecutionMs int64             `json:"execution_time_ms"`
+}
+
+// QueryGraphRAGRelationsActivity queries GraphRAG for relation patterns (stub)
+func QueryGraphRAGRelationsActivity(ctx context.Context, input QueryGraphRAGRelationsInput) (QueryGraphRAGRelationsOutput, error) {
+	output := QueryGraphRAGRelationsOutput{
+		Query:      input.Query,
+		Results:    []EdgeWithWording{},
+		TotalCount: 0,
+	}
+
+	// Stub implementation - actual GraphRAG querying would happen here
+	return output, nil
+}

activity/integration.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/judge.go

@@ -0,0 +1,32 @@
+package activity
+
+import (
+	"context"
+
+	"github.com/rockliang/poimen/workflows/pkg/types"
+)
+
+type JudgeInput struct {
+	Config              types.OrchestratorConfig
+	Diff                string
+	IntegrationTestLogs string
+}
+
+type JudgeOutput struct {
+	Verdict  string
+	Critique string
+}
+
+func JudgeActivity(ctx context.Context, in JudgeInput) (JudgeOutput, error) {
+	response, err := CallRoleLLM(ctx, in.Config, "judge", map[string]any{
+		"SystemPrompt": in.Config.SystemPrompt,
+		"Diff":         in.Diff,
+		"TestResult":   in.IntegrationTestLogs,
+	})
+	if err != nil {
+		return JudgeOutput{}, err
+	}
+
+	// TODO: parse LLM response for verdict
+	return JudgeOutput{Verdict: "pass", Critique: response}, nil
+}

activity/lessons.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/llm/client.go

@@ -9,7 +9,7 @@ import (
 	"net/http"
 	"os"
 
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/pkg/types"
 )
 
 var (
@@ -54,9 +54,10 @@ func NewClient() (*OpenAIClient, error) {
 
 // MessageInput is the input to CreateMessage.
 type MessageInput struct {
-	Model        statemachine.ModelSpec
+	Model        types.ModelSpec
 	SystemPrompt string
 	Messages     []MessageParam
+	AuthToken    string // Optional JWT token for authenticated endpoints
 }
 
 // MessageParam represents a message parameter.
@@ -136,6 +137,11 @@ func (c *OpenAIClient) CreateMessage(ctx context.Context, in MessageInput) (stri
 
 	httpReq.Header.Set("Content-Type", "application/json")
 
+	// Add authentication header if token provided
+	if in.AuthToken != "" {
+		httpReq.Header.Set("Authorization", fmt.Sprintf("Bearer %s", in.AuthToken))
+	}
+
 	// Send request
 	resp, err := c.httpClient.Do(httpReq)
 	if err != nil {

activity/llm/client_test.go

@@ -4,7 +4,7 @@ import (
 	"context"
 	"testing"
 
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/pkg/types"
 )
 
 func TestNewClient(t *testing.T) {
@@ -70,7 +70,7 @@ func TestCreateMessageValidation(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			in := MessageInput{
-				Model: statemachine.ModelSpec{
+				Model: types.ModelSpec{
 					ModelID: tt.modelID,
 				},
 				SystemPrompt: "test",

activity/llm_helper.go

@@ -0,0 +1,47 @@
+package activity
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/rockliang/poimen/workflows/activity/llm"
+	"github.com/rockliang/poimen/workflows/pkg/types"
+	"github.com/rockliang/poimen/workflows/prompts"
+)
+
+// CallRoleLLM is the shared pattern for calling an LLM with a role-based prompt.
+// Used by planner, implementer, and judge activities (DRY extraction).
+func CallRoleLLM(ctx context.Context, config types.OrchestratorConfig, role string, vars map[string]any) (string, error) {
+	client, err := llm.NewClient()
+	if err != nil {
+		return "", fmt.Errorf("failed to create LLM client: %w", err)
+	}
+
+	spec, exists := config.RolePrompts[role]
+	if !exists {
+		return "", fmt.Errorf("%s role prompt not configured", role)
+	}
+
+	// Render template
+	var content string
+	if spec.RawTemplate != "" {
+		content = spec.RawTemplate
+	} else {
+		content, err = prompts.Render(spec.TemplateRef, vars)
+		if err != nil {
+			return "", fmt.Errorf("failed to render %s template: %w", role, err)
+		}
+	}
+
+	// Call LLM
+	response, err := client.CreateMessage(ctx, llm.MessageInput{
+		Model:        spec.Model,
+		SystemPrompt: config.SystemPrompt,
+		Messages:     []llm.MessageParam{{Role: "user", Content: content}},
+	})
+	if err != nil {
+		return "", fmt.Errorf("%s LLM call failed: %w", role, err)
+	}
+
+	return response, nil
+}

activity/llm_inference.go

@@ -0,0 +1,114 @@
+package activity
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/rockliang/poimen/workflows/activity/llm"
+	"github.com/rockliang/poimen/workflows/pkg/types"
+)
+
+type LLMInferenceInput struct {
+	Model        string  `json:"model"`
+	SystemPrompt string  `json:"system_prompt"`
+	UserPrompt   string  `json:"user_prompt"`
+	Temperature  float64 `json:"temperature,omitempty"`
+	MaxTokens    int     `json:"max_tokens,omitempty"`
+	AuthToken    string  `json:"auth_token,omitempty"`
+}
+
+type LLMInferenceOutput struct {
+	Response     string `json:"response"`
+	Model        string `json:"model"`
+	StopReason   string `json:"stop_reason"`
+	TokensUsed   int    `json:"tokens_used"`
+	ErrorMessage string `json:"error,omitempty"`
+}
+
+func LLMInferenceActivity(ctx context.Context, in LLMInferenceInput) (LLMInferenceOutput, error) {
+	logger := newActivityLogger(ctx)
+	output := LLMInferenceOutput{Model: in.Model}
+
+	if in.Model == "" {
+		return output, fmt.Errorf("model not specified")
+	}
+	if in.UserPrompt == "" {
+		return output, fmt.Errorf("user_prompt not specified")
+	}
+
+	logger.Info("Starting LLM inference", "model", in.Model)
+
+	client, err := llm.NewClient()
+	if err != nil {
+		output.ErrorMessage = err.Error()
+		return output, fmt.Errorf("failed to create LLM client: %w", err)
+	}
+
+	response, err := client.CreateMessage(ctx, llm.MessageInput{
+		Model:        types.ModelSpec{ModelID: in.Model},
+		SystemPrompt: in.SystemPrompt,
+		Messages:     []llm.MessageParam{{Role: "user", Content: in.UserPrompt}},
+		AuthToken:    in.AuthToken,
+	})
+	if err != nil {
+		output.ErrorMessage = err.Error()
+		logger.Warn("LLM API call failed", "error", err)
+		return output, fmt.Errorf("LLM inference failed: %w", err)
+	}
+
+	output.Response = response
+	output.StopReason = "stop_sequence"
+	logger.Info("LLM inference completed", "response_len", len(response))
+	return output, nil
+}
+
+type LLMBatchInferenceInput struct {
+	Model        string   `json:"model"`
+	SystemPrompt string   `json:"system_prompt"`
+	Prompts      []string `json:"prompts"`
+	Temperature  float64  `json:"temperature,omitempty"`
+	AuthToken    string   `json:"auth_token,omitempty"`
+}
+
+type LLMBatchInferenceOutput struct {
+	Responses []string `json:"responses"`
+	Model     string   `json:"model"`
+	Errors    []string `json:"errors,omitempty"`
+}
+
+func LLMBatchInferenceActivity(ctx context.Context, in LLMBatchInferenceInput) (LLMBatchInferenceOutput, error) {
+	logger := newActivityLogger(ctx)
+	output := LLMBatchInferenceOutput{Model: in.Model, Responses: []string{}, Errors: []string{}}
+
+	if in.Model == "" {
+		return output, fmt.Errorf("model not specified")
+	}
+	if len(in.Prompts) == 0 {
+		return output, fmt.Errorf("no prompts provided")
+	}
+
+	logger.Info("Starting batch inference", "model", in.Model, "count", len(in.Prompts))
+
+	client, err := llm.NewClient()
+	if err != nil {
+		return output, fmt.Errorf("failed to create LLM client: %w", err)
+	}
+
+	for i, prompt := range in.Prompts {
+		response, err := client.CreateMessage(ctx, llm.MessageInput{
+			Model:        types.ModelSpec{ModelID: in.Model},
+			SystemPrompt: in.SystemPrompt,
+			Messages:     []llm.MessageParam{{Role: "user", Content: prompt}},
+		})
+		if err != nil {
+			output.Errors = append(output.Errors, fmt.Sprintf("prompt %d: %v", i, err))
+			output.Responses = append(output.Responses, "")
+			logger.Warn("Failed prompt", "index", i, "error", err)
+		} else {
+			output.Responses = append(output.Responses, response)
+		}
+	}
+
+	logger.Info("Batch inference completed", "responses", len(output.Responses), "errors", len(output.Errors))
+	return output, nil
+}

activity/logger.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/memory.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"
@@ -92,9 +92,9 @@ func RetrieveMemoryActivity(ctx context.Context, in RetrieveMemoryInput) (Retrie
 	}
 
 	// Get memory service URL and token
-	baseURL := os.Getenv("POIMEN_MEMORY_URL")
+	baseURL := os.Getenv("MEMORY_SERVICE_URL")
 	if baseURL == "" {
-		baseURL = "http://poimen-memory.poimen.svc.cluster.local:8080"
+		baseURL = "http://localhost:8080"
 	}
 
 	token := os.Getenv("POIMEN_MEMORY_TOKEN")

activity/memory_test.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"
@@ -39,7 +39,7 @@ func TestRetrieveMemoryActivity_Query(t *testing.T) {
 	defer server.Close()
 
 	// Set env for test
-	t.Setenv("POIMEN_MEMORY_URL", server.URL)
+	t.Setenv("MEMORY_SERVICE_URL", server.URL)
 
 	output, err := RetrieveMemoryActivity(context.Background(), RetrieveMemoryInput{
 		Query:   "security scanning",
@@ -93,7 +93,7 @@ func TestRetrieveMemoryActivity_Context(t *testing.T) {
 	}))
 	defer server.Close()
 
-	t.Setenv("POIMEN_MEMORY_URL", server.URL)
+	t.Setenv("MEMORY_SERVICE_URL", server.URL)
 
 	output, err := RetrieveMemoryActivity(context.Background(), RetrieveMemoryInput{
 		Query: "security scan repo",

activity/notification.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"bytes"

activity/notification_test.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/planner.go

@@ -0,0 +1,47 @@
+package activity
+
+import (
+	"context"
+
+	"github.com/rockliang/poimen/workflows/pkg/types"
+)
+
+type PlanningInput struct {
+	Config      types.OrchestratorConfig
+	BoardState  string
+	RepoPath    string
+	Milestone   string
+	TaskResults []types.TaskUnitOutput
+}
+
+type TaskDispatch struct {
+	TaskID      string
+	PromptSpec  types.PromptSpec
+	BaseTimeout *int64
+}
+
+type PlanningOutput struct {
+	TasksToDispatch      []string
+	CompletedBranches    []string
+	SubmilestoneComplete bool
+}
+
+func PlanningActivity(ctx context.Context, in PlanningInput) (PlanningOutput, error) {
+	response, err := CallRoleLLM(ctx, in.Config, "planner", map[string]any{
+		"SystemPrompt": in.Config.SystemPrompt,
+		"BoardState":   in.BoardState,
+		"Milestone":    in.Milestone,
+		"Config":       in.Config,
+	})
+	if err != nil {
+		return PlanningOutput{}, err
+	}
+
+	// TODO: parse LLM response into task dispatch list
+	_ = response
+	return PlanningOutput{
+		TasksToDispatch:      []string{},
+		CompletedBranches:    []string{},
+		SubmilestoneComplete: false,
+	}, nil
+}

activity/query_graph_rag.go

@@ -0,0 +1,101 @@
+package activity
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"time"
+)
+
+// QueryGraphRAGActivity queries Memory System for semantic relations
+func QueryGraphRAGActivity(ctx context.Context, input GraphRAGQueryInput) (GraphRAGQueryOutput, error) {
+	logger := newActivityLogger(ctx)
+	output := GraphRAGQueryOutput{
+		WorkflowID: input.WorkflowID,
+		Query:      input.Query,
+		Edges:      []EdgeWithWording{},
+		Paths:      []QueryPathData{},
+	}
+
+	logger.Info("Querying GraphRAG: %s", input.Query)
+
+	// Get Memory Service URL from env
+	memoryURL := os.Getenv("MEMORY_SERVICE_URL")
+	if memoryURL == "" {
+		memoryURL = "http://localhost:8000"
+	}
+
+	// Build payload for Memory System
+	payload := map[string]interface{}{
+		"workflow_id":      input.WorkflowID,
+		"query":            input.Query,
+		"search_type":      input.SearchType,
+		"relation_type":    input.RelationType,
+		"confidence_floor": input.ConfidenceFloor,
+		"top_k":            input.TopK,
+		"ranking_profile":  input.RankingProfile,
+		"canvas_nodes":     input.Canvas.Nodes,
+		"canvas_edges":     input.Canvas.Edges,
+		"relations":        input.Canvas.Relations,
+	}
+
+	reqBody, err := json.Marshal(payload)
+	if err != nil {
+		return output, fmt.Errorf("failed to marshal payload: %w", err)
+	}
+
+	// Call Memory System unified query endpoint
+	req, err := http.NewRequestWithContext(
+		ctx,
+		"POST",
+		memoryURL+"/workflows/query",
+		bytes.NewReader(reqBody),
+	)
+	if err != nil {
+		return output, fmt.Errorf("failed to create request: %w", err)
+	}
+
+	req.Header.Set("Content-Type", "application/json")
+	if token := ctx.Value("jwt_token"); token != nil {
+		req.Header.Set("Authorization", fmt.Sprintf("Bearer %v", token))
+	}
+
+	startTime := time.Now()
+	client := &http.Client{Timeout: 30 * time.Second}
+	resp, err := client.Do(req)
+	if err != nil {
+		return output, fmt.Errorf("failed to call Memory Service: %w", err)
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != 200 {
+		body, _ := io.ReadAll(resp.Body)
+		return output, fmt.Errorf("Memory Service returned %d: %s", resp.StatusCode, string(body))
+	}
+
+	// Parse response
+	var graphResp struct {
+		Edges      []EdgeWithWording `json:"edges"`
+		Paths      []QueryPathData   `json:"paths"`
+		TotalCount int               `json:"total_count"`
+		HasMore    bool              `json:"has_more"`
+	}
+
+	if err := json.NewDecoder(resp.Body).Decode(&graphResp); err != nil {
+		return output, fmt.Errorf("failed to decode response: %w", err)
+	}
+
+	output.Edges = graphResp.Edges
+	output.Paths = graphResp.Paths
+	output.TotalCount = graphResp.TotalCount
+	output.HasMore = graphResp.HasMore
+	output.ExecutionMs = time.Since(startTime).Milliseconds()
+
+	logger.Info("GraphRAG returned %d edges, %d paths in %dms", 
+		len(output.Edges), len(output.Paths), output.ExecutionMs)
+	return output, nil
+}

activity/router.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"

activity/skills.go

@@ -1,4 +1,4 @@
-package action
+package activity
 
 import (
 	"context"
@@ -10,12 +10,11 @@ import (
 	"time"
 
 	"go.temporal.io/sdk/activity"
-	"github.com/rockliang/poimen/workflows/statemachine"
 )
 
 // PrepareSkillsInput is input to PrepareSkillsActivity.
 type PrepareSkillsInput struct {
-	Skills        []statemachine.SkillRef
+	Skills        []SkillRef
 	StreamTimeout time.Duration
 	Provider      string // pi provider name (e.g. "homelab-reasoning"); required, pi has no usable default provider
 }

activity/types.go

@@ -0,0 +1,21 @@
+package activity
+
+import "github.com/rockliang/poimen/workflows/pkg/types"
+
+// Re-export from pkg/types for convenience within activity package.
+type ModelSpec = types.ModelSpec
+type PromptSpec = types.PromptSpec
+type SkillRef = types.SkillRef
+type OrchestratorConfig = types.OrchestratorConfig
+type TaskUnitOutput = types.TaskUnitOutput
+type EdgeWithWording = types.EdgeWithWording
+type RelationWording = types.RelationWording
+type CanvasWithRelationsData = types.CanvasWithRelationsData
+type FetchCanvasRelationsInput = types.FetchCanvasRelationsInput
+type CanvasReasonerInput = types.CanvasReasonerInput
+type GraphRAGQueryInput = types.GraphRAGQueryInput
+type GraphRAGQueryOutput = types.GraphRAGQueryOutput
+type QueryPathData = types.QueryPathData
+type CanvasCompatibilityInput = types.CanvasCompatibilityInput
+type IndexGraphRAGInput = types.IndexGraphRAGInput
+type IndexGraphRAGOutput = types.IndexGraphRAGOutput

cmd/starter/main.go

@@ -11,12 +11,12 @@ import (
 	"time"
 
 	"go.temporal.io/sdk/client"
-	"github.com/rockliang/poimen/workflows/action/llm"
+	"github.com/rockliang/poimen/workflows/activity/llm"
 	"github.com/rockliang/poimen/workflows/internal/config"
 	"github.com/rockliang/poimen/workflows/internal/health"
 	"github.com/rockliang/poimen/workflows/internal/logging"
 	"github.com/rockliang/poimen/workflows/internal/routing"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 )
 
 func main() {
@@ -89,20 +89,20 @@ func main() {
 
 
 	// Build OrchestratorInput
-	input := statemachine.OrchestratorInput{
+	input := workflow.OrchestratorInput{
 		TargetRepoPath:     *repoPath,
 		RemoteURL:          *remoteURL,
 		Milestone:          *milestone,
 		DryRun:             *dryRun,
 		MaxCyclesBeforeCAN: 100,
 		PiProvider:         *piProvider,
-		Config: statemachine.OrchestratorConfig{
+		Config: workflow.OrchestratorConfig{
 			SystemPrompt: "You are an expert software developer orchestrating multi-agent work.",
-			Skills:       []statemachine.SkillRef{},
-			RolePrompts: map[string]statemachine.PromptSpec{
+			Skills:       []workflow.SkillRef{},
+			RolePrompts: map[string]workflow.PromptSpec{
 				"planner": {
 					TemplateRef: "planner/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID:  *plannerModel,
 						Thinking: "adaptive",
 						Effort:   "high",
@@ -110,7 +110,7 @@ func main() {
 				},
 				"judge": {
 					TemplateRef: "judge/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID:  *judgeModel,
 						Thinking: "adaptive",
 						Effort:   "high",
@@ -118,12 +118,12 @@ func main() {
 				},
 				"implementer": {
 					TemplateRef: "implementer/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID: *implementerModel,
 					},
 				},
 			},
-			Tuning: statemachine.NewActivityTuning(),
+			Tuning: workflow.NewActivityTuning(),
 		},
 	}
 
@@ -144,7 +144,7 @@ func main() {
 	run, err := c.ExecuteWorkflow(context.Background(), client.StartWorkflowOptions{
 		ID:        workflowID,
 		TaskQueue: "poimen-taskqueue",
-	}, statemachine.OrchestratorWorkflow, input)
+	}, workflow.OrchestratorWorkflow, input)
 	if err != nil {
 		logging.Fatal("failed to start workflow", logging.Err(err))
 	}
@@ -164,7 +164,7 @@ func main() {
 	ctx, cancel := context.WithTimeout(context.Background(), 1*time.Minute)
 	defer cancel()
 
-	var result statemachine.OrchestratorOutput
+	var result workflow.OrchestratorOutput
 	if err := run.Get(ctx, &result); err != nil {
 		fmt.Printf("\nWorkflow initiated (execution in progress).\n")
 		fmt.Printf("Check the Web UI for real-time status updates.\n")
@@ -267,12 +267,12 @@ func runRoutingWorkflow(c client.Client, routeMsg, specFile string, isCron, dryR
 	}
 
 	workflowID := "routing-" + spec.Name + "-" + time.Now().Format("20060102-150405")
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
 	run, err := c.ExecuteWorkflow(ctx, client.StartWorkflowOptions{
 		ID:        workflowID,
 		TaskQueue: "poimen-taskqueue",
-	}, statemachine.RoutingWorkflow, input)
+	}, workflow.RoutingWorkflow, input)
 	if err != nil {
 		logging.Fatal("failed to start routing workflow", logging.Err(err))
 	}
@@ -285,7 +285,7 @@ func runRoutingWorkflow(c client.Client, routeMsg, specFile string, isCron, dryR
 	waitCtx, cancel := context.WithTimeout(ctx, 30*time.Second)
 	defer cancel()
 
-	var result statemachine.RoutingWorkflowOutput
+	var result workflow.RoutingWorkflowOutput
 	if err := run.Get(waitCtx, &result); err != nil {
 		fmt.Printf("\nWorkflow running (check Temporal UI for status)\n")
 	} else {

cmd/worker/main.go

@@ -11,11 +11,11 @@ import (
 
 	"go.temporal.io/sdk/client"
 	"go.temporal.io/sdk/worker"
-	"github.com/rockliang/poimen/workflows/action"
+	"github.com/rockliang/poimen/workflows/activity"
 	"github.com/rockliang/poimen/workflows/internal/config"
 	"github.com/rockliang/poimen/workflows/internal/health"
 	"github.com/rockliang/poimen/workflows/internal/logging"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 )
 
 func main() {
@@ -48,45 +48,56 @@ func main() {
 	}
 
 	// Register all workflows
-	w.RegisterWorkflow(statemachine.OrchestratorWorkflow)
-	w.RegisterWorkflow(statemachine.TaskUnitWorkflow)
-	w.RegisterWorkflow(statemachine.TestWorkflow)
-	w.RegisterWorkflow(statemachine.RoutingWorkflow)
+	w.RegisterWorkflow(workflow.OrchestratorWorkflow)
+	w.RegisterWorkflow(workflow.TaskUnitWorkflow)
+	w.RegisterWorkflow(workflow.TestWorkflow)
+	w.RegisterWorkflow(workflow.RoutingWorkflow)
+	w.RegisterWorkflow(workflow.WorkflowGraphQuery)
 
 	// Register all activities
-	w.RegisterActivity(action.CloneRepoActivity)
-	w.RegisterActivity(action.GitWorktreeAddActivity)
-	w.RegisterActivity(action.GitCommitActivity)
-	w.RegisterActivity(action.GitPushActivity)
-	w.RegisterActivity(action.GitSquashMergeActivity)
-	w.RegisterActivity(action.GitDiffActivity)
-	w.RegisterActivity(action.PrepareSkillsActivity)
-	w.RegisterActivity(action.PlanningActivity)
-	w.RegisterActivity(action.ImplementerActivity)
-	w.RegisterActivity(action.JudgeActivity)
+	w.RegisterActivity(activity.CloneRepoActivity)
+	w.RegisterActivity(activity.GitWorktreeAddActivity)
+	w.RegisterActivity(activity.GitCommitActivity)
+	w.RegisterActivity(activity.GitPushActivity)
+	w.RegisterActivity(activity.GitSquashMergeActivity)
+	w.RegisterActivity(activity.GitDiffActivity)
+	w.RegisterActivity(activity.PrepareSkillsActivity)
+	w.RegisterActivity(activity.PlanningActivity)
+	w.RegisterActivity(activity.ImplementerActivity)
+	w.RegisterActivity(activity.JudgeActivity)
 	// Integration and lessons activities - register when fully tested
-	w.RegisterActivity(action.RunIntegrationTestActivity)
-	// w.RegisterActivity(action.UpdateLessonsActivity)
-	// w.RegisterActivity(action.ReadLessonsActivity)
+	w.RegisterActivity(activity.RunIntegrationTestActivity)
+	// w.RegisterActivity(activity.UpdateLessonsActivity)
+	// w.RegisterActivity(activity.ReadLessonsActivity)
 
 	// Routing workflow activities
-	w.RegisterActivity(action.LLMRouterActivity)
-	w.RegisterActivity(action.ValidateWorkflowSpecActivity)
-	w.RegisterActivity(action.ValidateCronWorkflowSpecActivity)
+	w.RegisterActivity(activity.LLMRouterActivity)
+	w.RegisterActivity(activity.ValidateWorkflowSpecActivity)
+	w.RegisterActivity(activity.ValidateCronWorkflowSpecActivity)
 
 	// Analysis activities
-	w.RegisterActivity(action.AnalyzeCodeActivity)
-	w.RegisterActivity(action.SecurityScanActivity)
-	w.RegisterActivity(action.GenerateReportActivity)
+	w.RegisterActivity(activity.AnalyzeCodeActivity)
+	w.RegisterActivity(activity.SecurityScanActivity)
+	w.RegisterActivity(activity.GenerateReportActivity)
 
 	// Notification and utility activities
-	w.RegisterActivity(action.NotifyStatusActivity)
-	w.RegisterActivity(action.ArchiveResultsActivity)
-	w.RegisterActivity(action.DeploymentPreCheckActivity)
-	w.RegisterActivity(action.ApproveWorkflowActivity)
+	w.RegisterActivity(activity.NotifyStatusActivity)
+	w.RegisterActivity(activity.ArchiveResultsActivity)
+	w.RegisterActivity(activity.DeploymentPreCheckActivity)
+	w.RegisterActivity(activity.ApproveWorkflowActivity)
+
+	// Authentication activities
+	w.RegisterActivity(activity.AssumeRoleActivity)
 
 	// Memory activities
-	w.RegisterActivity(action.RetrieveMemoryActivity)
+	w.RegisterActivity(activity.RetrieveMemoryActivity)
+
+	// GraphRAG activities
+	w.RegisterActivity(activity.FetchCanvasRelationsActivity)
+	w.RegisterActivity(activity.QueryGraphRAGActivity)
+	w.RegisterActivity(activity.CanvasReasonerActivity)
+	w.RegisterActivity(activity.IndexGraphRAGActivity)
+	w.RegisterActivity(activity.CanvasCompatibilityActivity)
 
 	// Initialize health checker
 	healthChecker := health.NewChecker(c)
@@ -116,7 +127,7 @@ func main() {
 	// Run worker in a goroutine
 	workerErrChan := make(chan error, 1)
 	go func() {
-		logging.Info("starting worker on queue", logging.String("queue", "poimen-taskqueue"))
+		logging.Info("starting worker", logging.String("queue", "poimen"))
 		if err := w.Run(worker.InterruptCh()); err != nil {
 			workerErrChan <- err
 		}

go.mod

@@ -1,8 +1,10 @@
 module github.com/rockliang/poimen/workflows
 
-go 1.25.4
+go 1.26.0
 
 require (
+	github.com/google/uuid v1.6.0
+	github.com/lib/pq v1.12.3
 	github.com/prometheus/client_golang v1.24.1
 	github.com/stretchr/testify v1.12.1
 	go.temporal.io/sdk v1.48.0
@@ -16,10 +18,8 @@ require (
 	github.com/facebookgo/clock v0.0.0-20150410010913-600d898af40a // indirect
 	github.com/gogo/protobuf v1.3.2 // indirect
 	github.com/golang/mock v1.6.0 // indirect
-	github.com/google/uuid v1.6.0 // indirect
 	github.com/grpc-ecosystem/go-grpc-middleware/v2 v2.3.2 // indirect
 	github.com/grpc-ecosystem/grpc-gateway/v2 v2.22.0 // indirect
-	github.com/inconshreveable/mousetrap v1.1.0 // indirect
 	github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 // indirect
 	github.com/nexus-rpc/nexus-proto-annotations v0.1.0 // indirect
 	github.com/nexus-rpc/sdk-go v0.7.0 // indirect
@@ -27,8 +27,6 @@ require (
 	github.com/prometheus/common v0.70.1 // indirect
 	github.com/prometheus/procfs v0.21.1 // indirect
 	github.com/robfig/cron v1.2.0 // indirect
-	github.com/spf13/cobra v1.10.2 // indirect
-	github.com/spf13/pflag v1.0.9 // indirect
 	github.com/stretchr/objx v0.5.3 // indirect
 	go.temporal.io/api v1.63.4 // indirect
 	go.uber.org/multierr v1.11.0 // indirect

go.sum

@@ -2,7 +2,6 @@ github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM=
 github.com/beorn7/perks v1.0.1/go.mod h1:G2ZrVWU2WbWT9wwq4/hrbKbnv/1ERSJQ0ibhJ6rlkpw=
 github.com/cespare/xxhash/v2 v2.3.0 h1:UL815xU9SqsFlibzuggzjXhog7bL6oX9BbNZnL2UFvs=
 github.com/cespare/xxhash/v2 v2.3.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
-github.com/cpuguy83/go-md2man/v2 v2.0.6/go.mod h1:oOW0eioCTA6cOiMLiUPZOpcVxMig6NIQQ7OS05n1F4g=
 github.com/facebookgo/clock v0.0.0-20150410010913-600d898af40a h1:yDWHCSQ40h88yih2JAcL6Ls/kVkSE8GFACTGVnMPruw=
 github.com/facebookgo/clock v0.0.0-20150410010913-600d898af40a/go.mod h1:7Ga40egUymuWXxAe151lTNnCv97MddSOVsjpPPkityA=
 github.com/go-logr/logr v1.4.3 h1:CjnDlHq8ikf6E492q6eKboGOC0T8CDaOvkHCIg8idEI=
@@ -23,8 +22,6 @@ github.com/grpc-ecosystem/go-grpc-middleware/v2 v2.3.2 h1:sGm2vDRFUrQJO/Veii4h4z
 github.com/grpc-ecosystem/go-grpc-middleware/v2 v2.3.2/go.mod h1:wd1YpapPLivG6nQgbf7ZkG1hhSOXDhhn4MLTknx2aAc=
 github.com/grpc-ecosystem/grpc-gateway/v2 v2.22.0 h1:asbCHRVmodnJTuQ3qamDwqVOIjwqUPTYmYuemVOx+Ys=
 github.com/grpc-ecosystem/grpc-gateway/v2 v2.22.0/go.mod h1:ggCgvZ2r7uOoQjOyu2Y1NhHmEPPzzuhWgcza5M1Ji1I=
-github.com/inconshreveable/mousetrap v1.1.0 h1:wN+x4NVGpMsO7ErUn/mUI3vEoE6Jt13X2s0bqwp9tc8=
-github.com/inconshreveable/mousetrap v1.1.0/go.mod h1:vpF70FUmC8bwa3OWnCshd2FqLfsEA9PFc4w1p2J65bw=
 github.com/kisielk/errcheck v1.5.0/go.mod h1:pFxgyoBC7bSaBwPgfKdkLd5X25qrDl4LWUI2bnpBCr8=
 github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck=
 github.com/klauspost/compress v1.19.1 h1:VsB4HPswih7mmZ8WleSFQ75c/Ui1M4trX5oAsJnhSlk=
@@ -35,6 +32,8 @@ github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/kylelemons/godebug v1.1.0 h1:RPNrshWIDI6G2gRW9EHilWtl7Z6Sb1BR0xunSBf0SNc=
 github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
+github.com/lib/pq v1.12.3 h1:tTWxr2YLKwIvK90ZXEw8GP7UFHtcbTtty8zsI+YjrfQ=
+github.com/lib/pq v1.12.3/go.mod h1:/p+8NSbOcwzAEI7wiMXFlgydTwcgTr3OSKMsD2BitpA=
 github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 h1:C3w9PqII01/Oq1c1nUAm88MOHcQC9l5mIlSMApZMrHA=
 github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822/go.mod h1:+n7T8mK8HuQTcFwEeznm/DIxMOiR9yIdICNftLE1DvQ=
 github.com/nexus-rpc/nexus-proto-annotations v0.1.0 h1:2fELd+9sqUtNu6Fg//pw8YFsxOvp8vZ8hfP0nHhNI80=
@@ -53,11 +52,6 @@ github.com/robfig/cron v1.2.0 h1:ZjScXvvxeQ63Dbyxy76Fj3AT3Ut0aKsyd2/tl3DTMuQ=
 github.com/robfig/cron v1.2.0/go.mod h1:JGuDeoQd7Z6yL4zQhZ3OPEVHB7fL6Ka6skscFHfmt2k=
 github.com/rogpeppe/go-internal v1.11.0 h1:cWPaGQEPrBb5/AsnsZesgZZ9yb1OQ+GOISoDNXVBh4M=
 github.com/rogpeppe/go-internal v1.11.0/go.mod h1:ddIwULY96R17DhadqLgMfk9H9tvdUzkipdSkR5nkCZA=
-github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
-github.com/spf13/cobra v1.10.2 h1:DMTTonx5m65Ic0GOoRY2c16WCbHxOOw6xxezuLaBpcU=
-github.com/spf13/cobra v1.10.2/go.mod h1:7C1pvHqHw5A4vrJfjNwvOdzYu0Gml16OCs2GRiTUUS4=
-github.com/spf13/pflag v1.0.9 h1:9exaQaMOCwffKiiiYk6/BndUBv+iRViNW+4lEMi0PvY=
-github.com/spf13/pflag v1.0.9/go.mod h1:McXfInJRrz4CZXVZOBLb0bTZqETkiAhM9Iw0y3An2Bg=
 github.com/stretchr/objx v0.5.3 h1:jmXUvGomnU1o3W/V5h2VEradbpJDwGrzugQQvL0POH4=
 github.com/stretchr/objx v0.5.3/go.mod h1:rDQraq+vQZU7Fde9LOZLr8Tax6zZvy4kuNKF+QYS+U0=
 github.com/stretchr/testify v1.12.1 h1:EuwCh5fleGS7H32xRwO3wRGT7DxrDhLAT6FF8MpWDWE=
@@ -89,7 +83,6 @@ go.uber.org/zap v1.28.0 h1:IZzaP1Fv73/T/pBMLk4VutPl36uNC+OSUh3JLG3FIjo=
 go.uber.org/zap v1.28.0/go.mod h1:rDLpOi171uODNm/mxFcuYWxDsqWSAVkFdX4XojSKg/Q=
 go.yaml.in/yaml/v2 v2.4.4 h1:tuyd0P+2Ont/d6e2rl3be67goVK4R6deVxCUX5vyPaQ=
 go.yaml.in/yaml/v2 v2.4.4/go.mod h1:gMZqIpDtDqOfM0uNfy0SkpRhvUryYH0Z6wdMYcacYXQ=
-go.yaml.in/yaml/v3 v3.0.4/go.mod h1:DhzuOOF2ATzADvBadXxruRBLzYTpT36CKvDb3+aBEFg=
 go.yaml.in/yaml/v3 v3.0.5 h1:N6y/pJk8buWs9NY5ERU2HSMfm+IuD/OtfdAnq6kESPw=
 go.yaml.in/yaml/v3 v3.0.5/go.mod h1:HVTZu1O7/Vkt2N+BFy8Zza+lnLsABggaTM2ZpNIGuKg=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=

internal/routing/activity_knowledge_base.json

@@ -407,11 +407,265 @@
         "dependencies": [],
         "notes": "Network-dependent. First activity to run for context-aware routing. Fast timeout."
       }
+    },
+    {
+      "name": "AssumeRoleActivity",
+      "description": "Request temporary JWT token for accessing LLM APIs (like AWS AssumeRole)",
+      "category": "authentication",
+      "inputs": {
+        "identity": {
+          "type": "string",
+          "description": "User/service identity requesting access",
+          "required": true,
+          "examples": ["[email protected]", "service:poimen-worker"]
+        },
+        "clientId": {
+          "type": "string",
+          "description": "OAuth2 client ID (from vault if not provided)",
+          "required": false
+        },
+        "clientSecret": {
+          "type": "string",
+          "description": "OAuth2 client secret (from vault if not provided)",
+          "required": false
+        },
+        "scope": {
+          "type": "string",
+          "description": "Scope of access (e.g., 'llm:read' or 'llm:read llm:write')",
+          "required": true,
+          "examples": ["llm:read", "llm:read llm:write", "llm:admin"]
+        },
+        "durationSeconds": {
+          "type": "integer",
+          "description": "Token validity duration in seconds (default: 3600, max: 86400)",
+          "required": false,
+          "default": 3600
+        },
+        "authServerUrl": {
+          "type": "string",
+          "description": "Auth server URL (from AUTH_SERVER_URL env if not provided)",
+          "required": false
+        }
+      },
+      "outputs": {
+        "token": {
+          "type": "string",
+          "description": "JWT token for calling api.riotpiao.com"
+        },
+        "expiresAt": {
+          "type": "integer",
+          "description": "Token expiration time (Unix timestamp)"
+        },
+        "expiresIn": {
+          "type": "integer",
+          "description": "Seconds until token expires"
+        },
+        "tokenType": {
+          "type": "string",
+          "description": "Token type (typically 'Bearer')"
+        }
+      },
+      "constraints": {
+        "defaultTimeout": "30s",
+        "isFlaky": false,
+        "recommendedRetries": 2,
+        "retryBackoff": 1.5,
+        "dependencies": [],
+        "notes": "Must run before LLM Router to provide auth token. Call early in workflow."
+      }
+    },
+    {
+      "name": "LLMInferenceActivity",
+      "description": "Call LLM API with custom prompt and get response text",
+      "category": "llm",
+      "inputs": {
+        "model": {
+          "type": "string",
+          "description": "Model ID (reasoning, ornith:35b, ornith:13b, qwen2.5:3b)",
+          "required": true,
+          "examples": ["reasoning", "ornith:35b"]
+        },
+        "system_prompt": {
+          "type": "string",
+          "description": "System instruction for the model",
+          "required": false,
+          "default": ""
+        },
+        "user_prompt": {
+          "type": "string",
+          "description": "User message to send to the model",
+          "required": true
+        },
+        "temperature": {
+          "type": "number",
+          "description": "Sampling temperature (0.0-1.0, higher=more creative)",
+          "required": false,
+          "default": 0.7
+        },
+        "max_tokens": {
+          "type": "integer",
+          "description": "Maximum tokens in response",
+          "required": false
+        }
+      },
+      "outputs": {
+        "response": {
+          "type": "string",
+          "description": "LLM response text"
+        },
+        "model": {
+          "type": "string",
+          "description": "Model used for inference"
+        },
+        "stop_reason": {
+          "type": "string",
+          "description": "Why inference stopped (stop_sequence, length, etc)"
+        },
+        "tokens_used": {
+          "type": "integer",
+          "description": "Total tokens consumed"
+        }
+      },
+      "constraints": {
+        "defaultTimeout": "120s",
+        "isFlaky": true,
+        "recommendedRetries": 2,
+        "retryBackoff": 2.0,
+        "dependencies": [],
+        "notes": "API-dependent. Network flaky. Use for single prompts. See LLMBatchInferenceActivity for multiple."
+      }
+    },
+    {
+      "name": "LLMBatchInferenceActivity",
+      "description": "Call LLM API multiple times sequentially with different prompts",
+      "category": "llm",
+      "inputs": {
+        "model": {
+          "type": "string",
+          "description": "Model ID (reasoning, ornith:35b, ornith:13b, qwen2.5:3b)",
+          "required": true
+        },
+        "system_prompt": {
+          "type": "string",
+          "description": "System instruction (same for all prompts)",
+          "required": false
+        },
+        "prompts": {
+          "type": "array",
+          "description": "List of user prompts to process",
+          "required": true,
+          "items": {
+            "type": "string"
+          }
+        },
+        "temperature": {
+          "type": "number",
+          "description": "Sampling temperature (0.0-1.0)",
+          "required": false,
+          "default": 0.7
+        }
+      },
+      "outputs": {
+        "responses": {
+          "type": "array",
+          "description": "List of LLM responses (parallel to input prompts)",
+          "items": {
+            "type": "string"
+          }
+        },
+        "model": {
+          "type": "string",
+          "description": "Model used"
+        },
+        "errors": {
+          "type": "array",
+          "description": "Error messages for failed prompts",
+          "items": {
+            "type": "string"
+          }
+        }
+      },
+      "constraints": {
+        "defaultTimeout": "600s",
+        "isFlaky": true,
+        "recommendedRetries": 1,
+        "retryBackoff": 2.0,
+        "dependencies": [],
+        "notes": "Sequential processing of multiple prompts. Use for batch analysis, summarization, etc."
+      }
+    },
+    {
+      "name": "CanvasReasonerActivity",
+      "description": "Use LLM reasoning to infer and suggest connections between workflow activities",
+      "category": "workflow",
+      "inputs": {
+        "nodes": {
+          "type": "array",
+          "description": "Canvas workflow nodes to analyze",
+          "required": true,
+          "items": {
+            "type": "object",
+            "properties": {
+              "id": {"type": "string"},
+              "type": {"type": "string"},
+              "label": {"type": "string"}
+            }
+          }
+        },
+        "edges": {
+          "type": "array",
+          "description": "Existing edges in the workflow",
+          "required": false,
+          "items": {
+            "type": "object"
+          }
+        },
+        "preserve_existing": {
+          "type": "boolean",
+          "description": "If true, only suggest new edges; if false, redesign entire workflow",
+          "required": false,
+          "default": true
+        },
+        "auth_token": {
+          "type": "string",
+          "description": "JWT token for authenticated LLM calls",
+          "required": false
+        }
+      },
+      "outputs": {
+        "suggested_edges": {
+          "type": "array",
+          "description": "Edges suggested by LLM reasoning",
+          "items": {
+            "type": "object",
+            "properties": {
+              "source": {"type": "string"},
+              "target": {"type": "string"}
+            }
+          }
+        },
+        "reasoning": {
+          "type": "string",
+          "description": "LLM explanation of suggested connections"
+        },
+        "confidence": {
+          "type": "number",
+          "description": "Confidence score (0.0-1.0) of the suggestions"
+        }
+      },
+      "constraints": {
+        "defaultTimeout": "120s",
+        "isFlaky": true,
+        "recommendedRetries": 2,
+        "retryBackoff": 2.0,
+        "dependencies": [],
+        "notes": "Uses reasoning model to analyze workflow logic. Good for understanding data flow and connections between activities."
+      }
     }
   ],
   "metadata": {
-    "totalActivities": 9,
-    "lastUpdated": "2025-08-31T00:00:00Z",
+    "totalActivities": 13,
+    "lastUpdated": "2025-09-05T00:00:00Z",
     "categories": {
       "repository": 1,
       "analysis": 1,
@@ -421,7 +675,10 @@
       "notification": 1,
       "approval": 1,
       "storage": 1,
-      "memory": 1
+      "memory": 1,
+      "authentication": 1,
+      "llm": 2,
+      "workflow": 1
     }
   }
 }

internal/routing/canvas_converter.go

@@ -0,0 +1,179 @@
+package routing
+
+import (
+	"fmt"
+
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// CanvasConverter converts visual canvas to executable WorkflowSpec
+type CanvasConverter struct {
+	validator *CanvasValidator
+}
+
+// NewCanvasConverter creates a converter
+func NewCanvasConverter() *CanvasConverter {
+	return &CanvasConverter{
+		validator: NewCanvasValidator(),
+	}
+}
+
+// CanvasToWorkflowSpec converts canvas to WorkflowSpec
+func (cc *CanvasConverter) CanvasToWorkflowSpec(canvas *db.Canvas) (*WorkflowSpec, error) {
+	// Validate first
+	if err := cc.validator.ValidateCanvas(canvas); err != nil {
+		return nil, fmt.Errorf("canvas validation failed: %w", err)
+	}
+
+	// Get topological order
+	sortedNodes, err := cc.validator.TopoSort(canvas.Nodes, canvas.Edges)
+	if err != nil {
+		return nil, fmt.Errorf("topological sort failed: %w", err)
+	}
+
+	// Build states from sorted nodes
+	states := []State{}
+	nodeToState := make(map[string]int) // node ID to state index
+
+	for i, node := range sortedNodes {
+		state := cc.nodeToState(node, canvas.Edges)
+		states = append(states, state)
+		nodeToState[node.ID] = i
+	}
+
+	// Wire up transitions
+	for i, node := range sortedNodes {
+		outgoing := cc.getOutgoingEdges(node.ID, canvas.Edges)
+
+		if len(outgoing) == 0 {
+			// Last state - no transitions
+			continue
+		}
+
+		if len(outgoing) == 1 {
+			// Single outgoing edge
+			targetNode := outgoing[0]
+			targetIdx := nodeToState[targetNode]
+			if targetIdx > i {
+				states[i].Next = states[targetIdx].Name
+			}
+		} else {
+			// Multiple outgoing edges - parallel
+			states[i].Type = "Parallel"
+			branches := []interface{}{}
+			for _, targetNode := range outgoing {
+				branches = append(branches, map[string]string{
+					"state": states[nodeToState[targetNode]].Name,
+				})
+			}
+			if states[i].Branches == nil {
+				states[i].Branches = branches
+			}
+		}
+	}
+
+	spec := &WorkflowSpec{
+		Name:   canvas.Name,
+		Input:  map[string]interface{}{},
+		States: states,
+	}
+
+	return spec, nil
+}
+
+// nodeToState converts a canvas node to a workflow state
+func (cc *CanvasConverter) nodeToState(node db.WorkflowNode, edges []db.WorkflowEdge) State {
+	// Map node type to activity name
+	activityName := cc.mapActivityType(node.Type)
+
+	state := State{
+		Name:       node.ID,
+		Type:       TaskActivity,
+		Activity:   activityName,
+		Retry:      &RetryPolicy{MaxAttempts: 3, BackoffSeconds: 2},
+		Timeout:    "300s",
+		Parameters: node.Data,
+	}
+
+	return state
+}
+
+// mapActivityType maps canvas activity type to Poimen activity
+func (cc *CanvasConverter) mapActivityType(canvasType string) string {
+	typeMap := map[string]string{
+		"clone-repo":          "CloneRepoActivity",
+		"analyze-code":        "AnalyzeCodeActivity",
+		"security-scan":       "SecurityScanActivity",
+		"generate-report":     "GenerateReportActivity",
+		"deployment-precheck": "DeploymentPreCheckActivity",
+		"notify-status":       "NotifyStatusActivity",
+		"approve-workflow":    "ApproveWorkflowActivity",
+		"archive-results":     "ArchiveResultsActivity",
+		"retrieve-memory":     "RetrieveMemoryActivity",
+		"assume-role":         "AssumeRoleActivity",
+		"llm-inference":       "LLMInferenceActivity",
+		"llm-batch-inference": "LLMBatchInferenceActivity",
+		"canvas-reasoner":     "CanvasReasonerActivity",
+	}
+
+	if mapped, ok := typeMap[canvasType]; ok {
+		return mapped
+	}
+
+	return canvasType // fallback to type as-is
+}
+
+// getOutgoingEdges returns target node IDs for a given source node
+func (cc *CanvasConverter) getOutgoingEdges(nodeID string, edges []db.WorkflowEdge) []string {
+	targets := []string{}
+	seen := make(map[string]bool)
+
+	for _, edge := range edges {
+		if edge.Source == nodeID && !seen[edge.Target] {
+			targets = append(targets, edge.Target)
+			seen[edge.Target] = true
+		}
+	}
+
+	return targets
+}
+
+// CanvasToExecutionPlan converts canvas to sequential activity list
+func (cc *CanvasConverter) CanvasToExecutionPlan(canvas *db.Canvas) ([]ExecutionStep, error) {
+	// Validate first
+	if err := cc.validator.ValidateCanvas(canvas); err != nil {
+		return nil, fmt.Errorf("canvas validation failed: %w", err)
+	}
+
+	// Get topological order
+	sortedNodes, err := cc.validator.TopoSort(canvas.Nodes, canvas.Edges)
+	if err != nil {
+		return nil, fmt.Errorf("topological sort failed: %w", err)
+	}
+
+	steps := []ExecutionStep{}
+	for i, node := range sortedNodes {
+		step := ExecutionStep{
+			Index:        i,
+			NodeID:       node.ID,
+			ActivityName: cc.mapActivityType(node.Type),
+			Label:        node.Label,
+			Parameters:   node.Data,
+			Timeout:      "300s",
+		}
+		steps = append(steps, step)
+	}
+
+	return steps, nil
+}
+
+// ExecutionStep represents one activity in execution plan
+type ExecutionStep struct {
+	Index        int                   `json:"index"`
+	NodeID       string                `json:"node_id"`
+	ActivityName string                `json:"activity_name"`
+	Label        string                `json:"label"`
+	Parameters   map[string]interface{} `json:"parameters"`
+	Timeout      string                `json:"timeout"`
+	DependsOn    []int                 `json:"depends_on,omitempty"` // Indices of predecessor steps
+}

internal/routing/canvas_validator.go

@@ -0,0 +1,317 @@
+package routing
+
+import (
+	"fmt"
+	"strings"
+
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// CanvasValidator validates React Flow canvas (nodes + edges)
+type CanvasValidator struct {
+	activityRegistry map[string]bool
+}
+
+// NewCanvasValidator creates validator with activity registry
+func NewCanvasValidator() *CanvasValidator {
+	return &CanvasValidator{
+		activityRegistry: map[string]bool{
+			"clone-repo":              true,
+			"analyze-code":            true,
+			"security-scan":           true,
+			"generate-report":         true,
+			"deployment-precheck":     true,
+			"notify-status":           true,
+			"approve-workflow":        true,
+			"archive-results":         true,
+			"retrieve-memory":         true,
+			"assume-role":             true,
+			"llm-inference":           true,
+			"llm-batch-inference":     true,
+			"canvas-reasoner":         true,
+		},
+	}
+}
+
+// ValidateCanvas checks canvas structure, connectivity, and DAG
+func (cv *CanvasValidator) ValidateCanvas(canvas *db.Canvas) error {
+	if canvas == nil {
+		return fmt.Errorf("canvas is nil")
+	}
+
+	if len(canvas.Nodes) == 0 {
+		return fmt.Errorf("canvas has no nodes")
+	}
+
+	// Step 1: Validate nodes
+	if err := cv.validateNodes(canvas.Nodes); err != nil {
+		return fmt.Errorf("node validation failed: %w", err)
+	}
+
+	// Step 2: Validate edges
+	if err := cv.validateEdges(canvas.Nodes, canvas.Edges); err != nil {
+		return fmt.Errorf("edge validation failed: %w", err)
+	}
+
+	// Step 3: Check for cycles (must be DAG)
+	if err := cv.detectCycles(canvas.Nodes, canvas.Edges); err != nil {
+		return fmt.Errorf("cycle detected: %w", err)
+	}
+
+	// Step 4: Check connectivity (all nodes reachable from start)
+	if err := cv.validateConnectivity(canvas.Nodes, canvas.Edges); err != nil {
+		return fmt.Errorf("connectivity check failed: %w", err)
+	}
+
+	return nil
+}
+
+// validateNodes checks each node has required fields and valid type
+func (cv *CanvasValidator) validateNodes(nodes []db.WorkflowNode) error {
+	if len(nodes) == 0 {
+		return fmt.Errorf("no nodes in canvas")
+	}
+
+	nodeIds := make(map[string]bool)
+
+	for i, node := range nodes {
+		// Check required fields
+		if node.ID == "" {
+			return fmt.Errorf("node[%d] has empty ID", i)
+		}
+
+		if nodeIds[node.ID] {
+			return fmt.Errorf("node[%d] has duplicate ID: %s", i, node.ID)
+		}
+		nodeIds[node.ID] = true
+
+		if node.Label == "" {
+			return fmt.Errorf("node[%d] (%s) has empty label", i, node.ID)
+		}
+
+		if node.Position == nil {
+			return fmt.Errorf("node[%d] (%s) has no position", i, node.ID)
+		}
+
+		// Check activity type (if present)
+		if node.Type != "" && !cv.activityRegistry[strings.ToLower(node.Type)] {
+			return fmt.Errorf("node[%d] (%s) has unknown activity type: %s", i, node.ID, node.Type)
+		}
+
+		// Check data structure
+		if node.Data == nil {
+			return fmt.Errorf("node[%d] (%s) has no data", i, node.ID)
+		}
+	}
+
+	return nil
+}
+
+// validateEdges checks edges reference valid nodes
+func (cv *CanvasValidator) validateEdges(nodes []db.WorkflowNode, edges []db.WorkflowEdge) error {
+	nodeIds := make(map[string]bool)
+	for _, node := range nodes {
+		nodeIds[node.ID] = true
+	}
+
+	for i, edge := range edges {
+		// Check required fields
+		if edge.Source == "" {
+			return fmt.Errorf("edge[%d] has empty source", i)
+		}
+
+		if edge.Target == "" {
+			return fmt.Errorf("edge[%d] has empty target", i)
+		}
+
+		// Check source node exists
+		if !nodeIds[edge.Source] {
+			return fmt.Errorf("edge[%d] references unknown source node: %s", i, edge.Source)
+		}
+
+		// Check target node exists
+		if !nodeIds[edge.Target] {
+			return fmt.Errorf("edge[%d] references unknown target node: %s", i, edge.Target)
+		}
+
+		// Check self-loops (discouraged but allow for now)
+		if edge.Source == edge.Target {
+			// Could warn here but not fail
+		}
+	}
+
+	return nil
+}
+
+// detectCycles checks for cycles in the DAG (must be acyclic)
+func (cv *CanvasValidator) detectCycles(nodes []db.WorkflowNode, edges []db.WorkflowEdge) error {
+	// Build adjacency list
+	graph := make(map[string][]string)
+	inDegree := make(map[string]int)
+
+	for _, node := range nodes {
+		graph[node.ID] = []string{}
+		inDegree[node.ID] = 0
+	}
+
+	for _, edge := range edges {
+		graph[edge.Source] = append(graph[edge.Source], edge.Target)
+		inDegree[edge.Target]++
+	}
+
+	// Kahn's algorithm: topological sort
+	queue := []string{}
+	for _, node := range nodes {
+		if inDegree[node.ID] == 0 {
+			queue = append(queue, node.ID)
+		}
+	}
+
+	processed := 0
+	for len(queue) > 0 {
+		// Dequeue
+		current := queue[0]
+		queue = queue[1:]
+		processed++
+
+		// Visit neighbors
+		for _, neighbor := range graph[current] {
+			inDegree[neighbor]--
+			if inDegree[neighbor] == 0 {
+				queue = append(queue, neighbor)
+			}
+		}
+	}
+
+	// If we didn't process all nodes, there's a cycle
+	if processed != len(nodes) {
+		return fmt.Errorf("graph has cycle (processed %d/%d nodes)", processed, len(nodes))
+	}
+
+	return nil
+}
+
+// validateConnectivity checks all nodes are reachable from start nodes
+func (cv *CanvasValidator) validateConnectivity(nodes []db.WorkflowNode, edges []db.WorkflowEdge) error {
+	if len(nodes) == 0 {
+		return nil
+	}
+
+	// Build adjacency list
+	graph := make(map[string][]string)
+	inDegree := make(map[string]int)
+
+	for _, node := range nodes {
+		graph[node.ID] = []string{}
+		inDegree[node.ID] = 0
+	}
+
+	for _, edge := range edges {
+		graph[edge.Source] = append(graph[edge.Source], edge.Target)
+		inDegree[edge.Target]++
+	}
+
+	// Find start nodes (in-degree 0)
+	startNodes := []string{}
+	for _, node := range nodes {
+		if inDegree[node.ID] == 0 {
+			startNodes = append(startNodes, node.ID)
+		}
+	}
+
+	if len(startNodes) == 0 {
+		return fmt.Errorf("no start nodes found (all nodes have incoming edges)")
+	}
+
+	// BFS from all start nodes
+	visited := make(map[string]bool)
+	queue := startNodes
+
+	for len(queue) > 0 {
+		// Dequeue
+		current := queue[0]
+		queue = queue[1:]
+
+		if visited[current] {
+			continue
+		}
+		visited[current] = true
+
+		// Visit neighbors
+		for _, neighbor := range graph[current] {
+			if !visited[neighbor] {
+				queue = append(queue, neighbor)
+			}
+		}
+	}
+
+	// Check all nodes were visited
+	if len(visited) != len(nodes) {
+		unreached := []string{}
+		for _, node := range nodes {
+			if !visited[node.ID] {
+				unreached = append(unreached, node.ID)
+			}
+		}
+		return fmt.Errorf("unreachable nodes: %v", unreached)
+	}
+
+	return nil
+}
+
+// TopoSort returns nodes in topological order (execution order)
+func (cv *CanvasValidator) TopoSort(nodes []db.WorkflowNode, edges []db.WorkflowEdge) ([]db.WorkflowNode, error) {
+	if len(nodes) == 0 {
+		return []db.WorkflowNode{}, nil
+	}
+
+	// Build adjacency list and in-degree map
+	graph := make(map[string][]string)
+	inDegree := make(map[string]int)
+	nodeMap := make(map[string]db.WorkflowNode)
+
+	for _, node := range nodes {
+		graph[node.ID] = []string{}
+		inDegree[node.ID] = 0
+		nodeMap[node.ID] = node
+	}
+
+	for _, edge := range edges {
+		graph[edge.Source] = append(graph[edge.Source], edge.Target)
+		inDegree[edge.Target]++
+	}
+
+	// Kahn's algorithm
+	queue := []string{}
+	for _, node := range nodes {
+		if inDegree[node.ID] == 0 {
+			queue = append(queue, node.ID)
+		}
+	}
+
+	result := []db.WorkflowNode{}
+	processed := make(map[string]bool)
+
+	for len(queue) > 0 {
+		// Dequeue
+		current := queue[0]
+		queue = queue[1:]
+
+		result = append(result, nodeMap[current])
+		processed[current] = true
+
+		// Visit neighbors
+		for _, neighbor := range graph[current] {
+			inDegree[neighbor]--
+			if inDegree[neighbor] == 0 {
+				queue = append(queue, neighbor)
+			}
+		}
+	}
+
+	if len(result) != len(nodes) {
+		return nil, fmt.Errorf("topological sort failed: graph has cycle")
+	}
+
+	return result, nil
+}

internal/routing/llm_client.go

@@ -8,7 +8,6 @@ import (
 	"io"
 	"net/http"
 	"os"
-	"strings"
 )
 
 var (
@@ -53,12 +52,6 @@ type LLMAuth struct {
 	
 	// HeaderValue is the custom header value for Custom auth
 	HeaderValue string `json:"headerValue,omitempty"`
-	
-	// TenantID is the tenant/customer ID for multi-tenant federated access
-	TenantID string `json:"tenantId,omitempty"`
-	
-	// Scopes are the OAuth2 scopes (space-separated)
-	Scopes string `json:"scopes,omitempty"`
 }
 
 // LLMClient is a simple LLM client for routing
@@ -216,16 +209,6 @@ func (c *LLMClient) applyAuth(req *http.Request) error {
 		req.Header.Set(c.auth.HeaderName, c.auth.HeaderValue)
 	}
 
-	// Add tenant ID if specified (for multi-tenant federated access)
-	if c.auth.TenantID != "" {
-		req.Header.Set("X-Tenant-ID", c.auth.TenantID)
-	}
-
-	// Add scopes if specified (for OAuth2 flows)
-	if c.auth.Scopes != "" {
-		req.Header.Set("X-OAuth-Scopes", c.auth.Scopes)
-	}
-
 	return nil
 }
 

internal/routing/llm_router.go

@@ -69,7 +69,6 @@ type LLMRouterConfig struct {
 	Validators     []WorkflowValidator
 	ParamBinder    ParameterBinder
 	Auth           *LLMAuth // Authentication config for LLM API
-	TenantID       string   // Tenant ID for multi-tenant isolation
 }
 
 // NewLLMRouter creates a new LLM router with custom config

internal/routing/types.go

@@ -28,12 +28,16 @@ type State struct {
 	Type              StateType `json:"type"`
 	
 	// Task fields
+	Activity          string                 `json:"activity,omitempty"`
 	Resource          string                 `json:"resource,omitempty"`
 	Parameters        map[string]interface{} `json:"parameters,omitempty"`
 	Timeout           string                 `json:"timeout,omitempty"`
 	Retry             *RetryPolicy           `json:"retry,omitempty"`
 	Catch             []CatchClause          `json:"catch,omitempty"`
 	
+	// Parallel fields
+	Branches          []interface{}          `json:"branches,omitempty"`
+	
 	// Pass fields
 	Result            interface{} `json:"result,omitempty"`
 	
@@ -50,14 +54,18 @@ type State struct {
 type StateType string
 
 const (
-	StateTypeTask StateType = "Task"
-	StateTypePass StateType = "Pass"
-	StateTypeFail StateType = "Fail"
+	StateTypeTask     StateType = "Task"
+	StateTypePass     StateType = "Pass"
+	StateTypeFail     StateType = "Fail"
+	StateTypeParallel StateType = "Parallel"
+
+	TaskActivity = "Task"
 )
 
 // RetryPolicy defines retry behavior for activities
 type RetryPolicy struct {
 	MaxAttempts     int32   `json:"maxAttempts"`
+	BackoffSeconds  int32   `json:"backoffSeconds,omitempty"`
 	BackoffRate     float64 `json:"backoffRate"`
 	InitialInterval string  `json:"initialInterval"`
 	MaxInterval     string  `json:"maxInterval,omitempty"`

k8s/configmap.enc.yaml

@@ -0,0 +1,27 @@
+apiVersion: ENC[AES256_GCM,data:IxE=,iv:nh5IQck87AsRYnvxMLxn2rZFBUTHcc9obvsYoHnvC6g=,tag:ink4m5kv1NqX2TeRrpCaqg==,type:str]
+kind: ENC[AES256_GCM,data:2U5oBCqywkR0,iv:EVBlp6G1SlznoP7Zx9Y0mQOnxbzcosP28+UbJUWFjHM=,tag:v1lot4q06QBqU2M5q2RJ/g==,type:str]
+metadata:
+    name: ENC[AES256_GCM,data:pFOLyxGAW7/KqUqijkPM1mGYaKo=,iv:jttNW2Ef0itpuWupOQkGEiFaanpcd/HuvfLg9l2tSEg=,tag:6ijOdYdjnUaCOf29z8vkxg==,type:str]
+    namespace: ENC[AES256_GCM,data:FSZFmVEp,iv:liL0yGuEbjZjf6egh8KS9zi6H3AEF9MvynGBO+49GW0=,tag:4iJNECe8vsXqzQBV/Q7c0g==,type:str]
+data:
+    #ENC[AES256_GCM,data:d+Z/w/n69JI5,iv:/w2xIrmNXg53kz+tfebT0P6rD3TC909YTDFXXUc/yHo=,tag:ao4yA2lW2a1/ELT8ZIBOgw==,type:comment]
+    TEMPORAL_NAMESPACE: ENC[AES256_GCM,data:X0D/DVfVgzhFh+oB8nU=,iv:biiLtXRALsUMNIW4qGj8JOc9C1RJK93pOw74ZmymM/g=,tag:tW5+AwUGj8jzvxPSZAf41A==,type:str]
+    TEMPORAL_HOSTPORT: ENC[AES256_GCM,data:OK8Cu7nA2fKhXmBko8MORoD6hvJBMFfE/dTBZ99ac420xjPBWjhrz2XR3WRiWhUIIQ==,iv:HKDNRT/30HNzN3pX91e63i/ZeIUUYkycknjjKTFlVwo=,tag:OjPle4p+nV4ZQYN/yuUFEw==,type:str]
+    #ENC[AES256_GCM,data:1jnEt80ScfSqnCv9Rf/XLAlYTfuZVxoRdV7Sx9EgIGmzvAM=,iv:3NKsB9ZytMxuMipSLApKX7hj2jQG+2e2Hp8+ofFin9k=,tag:s4eRatHTQJa37xBO5UYI5Q==,type:comment]
+    LOCAL_LLM_BASE_URL: ENC[AES256_GCM,data:rTWBhNovjp5WdFI2jMaMRRg41uhh1dOFqVnplmMADiubu/FGXCvUPD2d9IuU,iv:1RdS5UfD6aOXnaEfddRd4PxJ1+K7AwKlMUYxzUQxVJ4=,tag:K0g2KDCFLOtxDFkHxoRbMw==,type:str]
+    MEMORY_SERVICE_URL: ENC[AES256_GCM,data:CarE5ENFtDvBWVzB7qTZQqoL0OsN2uAO9aoPXu1LPWtdoSQ2rQsVs2scTA08jaWLD+o=,iv:e7M/aibmYdSk4NbY6fYw0UPjd/VQvWJ1tC8n0r2y6r0=,tag:bWuCTBRfKkaB2fauPR51Fg==,type:str]
+sops:
+    age:
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBBV0RJU2wzN1Bib0lZRlFl
+            K2UzSVBNUjNHUVVFeUFEWTJBUHNZT0l3RWprCmp2cE1HM2xOWnVjVGlqdWI0SndG
+            Q1VOaklVdlg4eGp6V29uVDVJRUtmKzgKLS0tIGp3MWJMdmdyZmRnWmxOWEhVZUhm
+            Rm5DNzBhUWt0SVBDK09mSWw2SHVRTWMKLS6Cz2nhz1RSoV+VUvFw9EsjlWbE2nEk
+            4P1FdNgr+v1MTNczKZGzh9HbTAmYxCRgVBszXR5ov2JfukaZAWb9mg==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1e5fq3hwxy78psus2nfvmtmua36g0u3suk78ephw6246l974d2utsvn0hla
+    lastmodified: "2026-09-06T12:56:33Z"
+    mac: ENC[AES256_GCM,data:Ra89XdYHoJV+uFBlUMdw+I36UQfhM+r0Q4WjI3V/A7jBY2t00I0xSP+KRKPXlAA18HXpLCc5VkWklPMU4gfyYied1bPuFw6yLvvuup/ev4UAg/LofThKjtuRfXPy82Ltvro6R+ilWzJ8pRtgtk+jC4xkkMsUOACJMVwgn979D7M=,iv:UWR+CqXAreW/fx3tt1kJsOKYZkrAkxLXlU5QVy5p3gE=,tag:3t9c6631D4XrDmkR6Iv/4Q==,type:str]
+    unencrypted_suffix: _unencrypted
+    version: 3.13.2

k8s/configmap.yaml

@@ -1,10 +0,0 @@
-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: poimen-config
-  namespace: poimen
-data:
-  TEMPORAL_NAMESPACE: "poimen-harness"
-  TEMPORAL_HOSTPORT: "temporal-frontend.temporal:7233"
-  LOCAL_LLM_BASE_URL: "http://api-gateway.api:8080"
-  POIMEN_MEMORY_URL: "http://poimen-memory.poimen.svc.cluster.local:8080"

k8s/git-commit.yaml

@@ -9,6 +9,6 @@ metadata:
     app.kubernetes.io/name: poimen
     app.kubernetes.io/component: orchestrator
 data:
-  GIT_COMMIT: "30644a8e"  # Updated automatically by CI/CD
+  GIT_COMMIT: "84b4ca120"  # Updated automatically by CI/CD
   GIT_BRANCH: "main"
-  DEPLOYMENT_DATE: "2026-09-04"
+  DEPLOYMENT_DATE: "2026-09-05"

k8s/kustomization.yaml

@@ -4,15 +4,19 @@ kind: Kustomization
 namespace: poimen
 
 resources:
-  - worker-deployment.yaml
-  - configmap.yaml
+  - poimen-application.yaml
 
 commonLabels:
   app.kubernetes.io/name: poimen
   app.kubernetes.io/component: worker
 
-secretGenerator:
-  - name: poimen-secrets
-    envs:
-      - secrets.env
-    behavior: create
+images:
+  - name: forgejo.riotpiao.com/rock/poimen-memory
+    newName: forgejo.riotpiao.com/rock/poimen-memory
+    newTag: latest
+  - name: forgejo.riotpiao.com/rock/poimen-workflows
+    newName: forgejo.riotpiao.com/rock/poimen-workflows
+    newTag: latest
+  - name: forgejo.riotpiao.com/rock/poimen-frontend
+    newName: forgejo.riotpiao.com/rock/poimen-frontend
+    newTag: latest

k8s/secret.yaml

@@ -1,12 +0,0 @@
-# NOTE: This file is for reference only.
-# Kustomize will auto-generate secrets from secrets.env
-# See kustomization.yaml for details
-
-apiVersion: v1
-kind: Secret
-metadata:
-  name: poimen-secrets
-  namespace: poimen
-type: Opaque
-stringData:
-  ANTHROPIC_API_KEY: "" # Generated from secrets.env by Kustomize

k8s/secrets.env

@@ -1 +0,0 @@
-ANTHROPIC_API_KEY=YOUR_ANTHROPIC_API_KEY_HERE

k8s/worker-deployment.yaml

@@ -13,8 +13,8 @@ spec:
       labels:
         app: poimen-worker
       annotations:
-        git-commit: "30644a8e"  # ✅ Updated on each push, triggers rolling restart
-        deployment-date: "2026-09-04"
+        git-commit: "84b4ca120"  # ✅ Updated on each push, triggers rolling restart
+        deployment-date: "2026-09-05"
     spec:
       containers:
       - name: worker

pkg/db/db.go

@@ -0,0 +1,535 @@
+package db
+
+import (
+	"context"
+	"database/sql"
+	"encoding/json"
+	"fmt"
+	"os"
+	"time"
+
+	_ "github.com/lib/pq"
+)
+
+// DB wraps the database connection
+type DB struct {
+	conn *sql.DB
+}
+
+// New creates a new database connection to memory-db (K8s CNPG)
+// Expected DSN format: postgresql://app:password@host:5432/dbname?sslmode=disable
+func New(dsn string) (*DB, error) {
+	if dsn == "" {
+		// Fallback: try to construct from K8s env vars
+		host := os.Getenv("DATABASE_HOST")
+		port := os.Getenv("DATABASE_PORT")
+		name := os.Getenv("DATABASE_NAME")
+		user := os.Getenv("DATABASE_USER")
+		password := os.Getenv("DATABASE_PASSWORD")
+		
+		if host != "" && port != "" && name != "" && user != "" && password != "" {
+			dsn = fmt.Sprintf("postgresql://%s:%s@%s:%s/%s?sslmode=disable", 
+				user, password, host, port, name)
+		} else {
+			return nil, fmt.Errorf("DATABASE_URL or K8s env vars (DATABASE_HOST, DATABASE_PORT, DATABASE_NAME, DATABASE_USER, DATABASE_PASSWORD) required")
+		}
+	}
+
+	conn, err := sql.Open("postgres", dsn)
+	if err != nil {
+		return nil, fmt.Errorf("failed to open database: %w", err)
+	}
+
+	// Test connection
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	if err := conn.PingContext(ctx); err != nil {
+		return nil, fmt.Errorf("failed to ping database: %w", err)
+	}
+
+	// Set connection pool settings
+	conn.SetMaxOpenConns(25)
+	conn.SetMaxIdleConns(5)
+	conn.SetConnMaxLifetime(5 * time.Minute)
+
+	return &DB{conn: conn}, nil
+}
+
+// Close closes the database connection
+func (db *DB) Close() error {
+	return db.conn.Close()
+}
+
+// SaveWorkflow saves or updates a workflow with canvas
+func (db *DB) SaveWorkflow(ctx context.Context, wf *Workflow) error {
+	query := `
+		INSERT INTO workflows (id, customer_id, name, description, status, version, nodes, edges, created_by, created_at, updated_at)
+		VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11)
+		ON CONFLICT (id) DO UPDATE SET
+			name = $3,
+			description = $4,
+			status = $5,
+			version = $6,
+			nodes = $7,
+			edges = $8,
+			updated_at = $11
+	`
+
+	_, err := db.conn.ExecContext(ctx, query,
+		wf.ID,
+		wf.CustomerID,
+		wf.Name,
+		wf.Description,
+		wf.Status,
+		wf.Version,
+		wf.Nodes,
+		wf.Edges,
+		wf.CreatedBy,
+		wf.CreatedAt,
+		wf.UpdatedAt,
+	)
+
+	return err
+}
+
+// SaveCanvasUpdate saves canvas (nodes + edges) for a workflow
+func (db *DB) SaveCanvasUpdate(ctx context.Context, workflowID, customerID string, canvas *Canvas) error {
+	nodesJSON, err := json.Marshal(canvas.Nodes)
+	if err != nil {
+		return fmt.Errorf("failed to marshal nodes: %w", err)
+	}
+
+	edgesJSON, err := json.Marshal(canvas.Edges)
+	if err != nil {
+		return fmt.Errorf("failed to marshal edges: %w", err)
+	}
+
+	query := `
+		UPDATE workflows
+		SET nodes = $1, edges = $2, updated_at = now()
+		WHERE id = $3 AND customer_id = $4
+	`
+
+	result, err := db.conn.ExecContext(ctx, query, nodesJSON, edgesJSON, workflowID, customerID)
+	if err != nil {
+		return fmt.Errorf("failed to update canvas: %w", err)
+	}
+
+	rows, err := result.RowsAffected()
+	if err != nil {
+		return err
+	}
+
+	if rows == 0 {
+		return fmt.Errorf("workflow not found: %s", workflowID)
+	}
+
+	return nil
+}
+
+// FetchWorkflow retrieves a workflow by ID
+func (db *DB) FetchWorkflow(ctx context.Context, workflowID, customerID string) (*Workflow, error) {
+	query := `
+		SELECT id, customer_id, name, description, status, version, nodes, edges, created_by, created_at, updated_at, last_executed_at
+		FROM workflows
+		WHERE id = $1 AND customer_id = $2
+	`
+
+	wf := &Workflow{}
+	err := db.conn.QueryRowContext(ctx, query, workflowID, customerID).Scan(
+		&wf.ID,
+		&wf.CustomerID,
+		&wf.Name,
+		&wf.Description,
+		&wf.Status,
+		&wf.Version,
+		&wf.Nodes,
+		&wf.Edges,
+		&wf.CreatedBy,
+		&wf.CreatedAt,
+		&wf.UpdatedAt,
+		&wf.LastExecutedAt,
+	)
+
+	if err != nil {
+		if err == sql.ErrNoRows {
+			return nil, fmt.Errorf("workflow not found: %s", workflowID)
+		}
+		return nil, fmt.Errorf("failed to fetch workflow: %w", err)
+	}
+
+	return wf, nil
+}
+
+// FetchCanvas retrieves canvas (nodes + edges) for a workflow
+func (db *DB) FetchCanvas(ctx context.Context, workflowID, customerID string) (*Canvas, error) {
+	query := `
+		SELECT nodes, edges
+		FROM workflows
+		WHERE id = $1 AND customer_id = $2
+	`
+
+	var nodesJSON, edgesJSON []byte
+	err := db.conn.QueryRowContext(ctx, query, workflowID, customerID).Scan(&nodesJSON, &edgesJSON)
+	if err != nil {
+		if err == sql.ErrNoRows {
+			return nil, fmt.Errorf("workflow not found: %s", workflowID)
+		}
+		return nil, fmt.Errorf("failed to fetch canvas: %w", err)
+	}
+
+	var nodes []WorkflowNode
+	var edges []WorkflowEdge
+
+	if err := json.Unmarshal(nodesJSON, &nodes); err != nil {
+		return nil, fmt.Errorf("failed to unmarshal nodes: %w", err)
+	}
+
+	if err := json.Unmarshal(edgesJSON, &edges); err != nil {
+		return nil, fmt.Errorf("failed to unmarshal edges: %w", err)
+	}
+
+	return &Canvas{Nodes: nodes, Edges: edges}, nil
+}
+
+// ListWorkflows retrieves all workflows for a customer
+func (db *DB) ListWorkflows(ctx context.Context, customerID string, limit, offset int) ([]Workflow, error) {
+	query := `
+		SELECT id, customer_id, name, description, status, version, nodes, edges, created_by, created_at, updated_at, last_executed_at
+		FROM workflows
+		WHERE customer_id = $1
+		ORDER BY created_at DESC
+		LIMIT $2 OFFSET $3
+	`
+
+	rows, err := db.conn.QueryContext(ctx, query, customerID, limit, offset)
+	if err != nil {
+		return nil, fmt.Errorf("failed to list workflows: %w", err)
+	}
+	defer rows.Close()
+
+	var workflows []Workflow
+	for rows.Next() {
+		wf := Workflow{}
+		err := rows.Scan(
+			&wf.ID,
+			&wf.CustomerID,
+			&wf.Name,
+			&wf.Description,
+			&wf.Status,
+			&wf.Version,
+			&wf.Nodes,
+			&wf.Edges,
+			&wf.CreatedBy,
+			&wf.CreatedAt,
+			&wf.UpdatedAt,
+			&wf.LastExecutedAt,
+		)
+		if err != nil {
+			return nil, fmt.Errorf("failed to scan workflow: %w", err)
+		}
+		workflows = append(workflows, wf)
+	}
+
+	return workflows, rows.Err()
+}
+
+// DeleteWorkflow deletes a workflow
+func (db *DB) DeleteWorkflow(ctx context.Context, workflowID, customerID string) error {
+	query := `
+		DELETE FROM workflows
+		WHERE id = $1 AND customer_id = $2
+	`
+
+	result, err := db.conn.ExecContext(ctx, query, workflowID, customerID)
+	if err != nil {
+		return fmt.Errorf("failed to delete workflow: %w", err)
+	}
+
+	rows, err := result.RowsAffected()
+	if err != nil {
+		return err
+	}
+
+	if rows == 0 {
+		return fmt.Errorf("workflow not found: %s", workflowID)
+	}
+
+	return nil
+}
+
+// SaveExecution saves a workflow execution record
+func (db *DB) SaveExecution(ctx context.Context, exec *WorkflowExecution) error {
+	query := `
+		INSERT INTO workflow_executions (id, workflow_id, customer_id, temporal_id, status, inputs, outputs, started_at, completed_at, duration_ms, error_message, error_count)
+		VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12)
+		ON CONFLICT (id) DO UPDATE SET
+			status = $5,
+			outputs = $7,
+			completed_at = $9,
+			duration_ms = $10,
+			error_message = $11,
+			error_count = $12
+	`
+
+	_, err := db.conn.ExecContext(ctx, query,
+		exec.ID,
+		exec.WorkflowID,
+		exec.CustomerID,
+		exec.TemporalID,
+		exec.Status,
+		exec.Inputs,
+		exec.Outputs,
+		exec.StartedAt,
+		exec.CompletedAt,
+		exec.DurationMs,
+		exec.ErrorMessage,
+		exec.ErrorCount,
+	)
+
+	return err
+}
+
+// FetchExecution retrieves a workflow execution
+func (db *DB) FetchExecution(ctx context.Context, executionID string) (*WorkflowExecution, error) {
+	query := `
+		SELECT id, workflow_id, customer_id, temporal_id, status, inputs, outputs, started_at, completed_at, duration_ms, error_message, error_count
+		FROM workflow_executions
+		WHERE id = $1
+	`
+
+	exec := &WorkflowExecution{}
+	err := db.conn.QueryRowContext(ctx, query, executionID).Scan(
+		&exec.ID,
+		&exec.WorkflowID,
+		&exec.CustomerID,
+		&exec.TemporalID,
+		&exec.Status,
+		&exec.Inputs,
+		&exec.Outputs,
+		&exec.StartedAt,
+		&exec.CompletedAt,
+		&exec.DurationMs,
+		&exec.ErrorMessage,
+		&exec.ErrorCount,
+	)
+
+	if err != nil {
+		if err == sql.ErrNoRows {
+			return nil, fmt.Errorf("execution not found: %s", executionID)
+		}
+		return nil, fmt.Errorf("failed to fetch execution: %w", err)
+	}
+
+	return exec, nil
+}
+
+// SaveExecutionLog saves an activity log entry
+func (db *DB) SaveExecutionLog(ctx context.Context, log *ExecutionLog) error {
+	query := `
+		INSERT INTO execution_logs (execution_id, node_id, activity_name, level, message, metadata, logged_at)
+		VALUES ($1, $2, $3, $4, $5, $6, $7)
+	`
+
+	_, err := db.conn.ExecContext(ctx, query,
+		log.ExecutionID,
+		log.NodeID,
+		log.ActivityName,
+		log.Level,
+		log.Message,
+		log.Metadata,
+		log.LoggedAt,
+	)
+
+	return err
+}
+
+// FetchExecutionLogs retrieves all logs for an execution
+func (db *DB) FetchExecutionLogs(ctx context.Context, executionID string) ([]ExecutionLog, error) {
+	query := `
+		SELECT id, execution_id, node_id, activity_name, level, message, metadata, logged_at
+		FROM execution_logs
+		WHERE execution_id = $1
+		ORDER BY logged_at ASC
+	`
+
+	rows, err := db.conn.QueryContext(ctx, query, executionID)
+	if err != nil {
+		return nil, fmt.Errorf("failed to fetch execution logs: %w", err)
+	}
+	defer rows.Close()
+
+	var logs []ExecutionLog
+	for rows.Next() {
+		log := ExecutionLog{}
+		err := rows.Scan(
+			&log.ID,
+			&log.ExecutionID,
+			&log.NodeID,
+			&log.ActivityName,
+			&log.Level,
+			&log.Message,
+			&log.Metadata,
+			&log.LoggedAt,
+		)
+		if err != nil {
+			return nil, fmt.Errorf("failed to scan log: %w", err)
+		}
+		logs = append(logs, log)
+	}
+
+	return logs, rows.Err()
+}
+
+// SaveActivityTrace saves per-activity execution trace
+func (db *DB) SaveActivityTrace(ctx context.Context, trace *ActivityTrace) error {
+	query := `
+		INSERT INTO activity_traces (execution_id, node_id, activity_name, parameters, result, started_at, completed_at, duration_ms, attempt, retry_reason, status, error_message)
+		VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12)
+		ON CONFLICT (id) DO UPDATE SET
+			status = $11,
+			result = $5,
+			completed_at = $7,
+			duration_ms = $8,
+			error_message = $12
+	`
+
+	_, err := db.conn.ExecContext(ctx, query,
+		trace.ExecutionID,
+		trace.NodeID,
+		trace.ActivityName,
+		trace.Parameters,
+		trace.Result,
+		trace.StartedAt,
+		trace.CompletedAt,
+		trace.DurationMs,
+		trace.Attempt,
+		trace.RetryReason,
+		trace.Status,
+		trace.ErrorMessage,
+	)
+
+	return err
+}
+
+// FetchActivityTraces retrieves all activity traces for an execution
+func (db *DB) FetchActivityTraces(ctx context.Context, executionID string) ([]ActivityTrace, error) {
+	query := `
+		SELECT id, execution_id, node_id, activity_name, parameters, result, started_at, completed_at, duration_ms, attempt, retry_reason, status, error_message
+		FROM activity_traces
+		WHERE execution_id = $1
+		ORDER BY started_at ASC
+	`
+
+	rows, err := db.conn.QueryContext(ctx, query, executionID)
+	if err != nil {
+		return nil, fmt.Errorf("failed to fetch activity traces: %w", err)
+	}
+	defer rows.Close()
+
+	var traces []ActivityTrace
+	for rows.Next() {
+		trace := ActivityTrace{}
+		err := rows.Scan(
+			&trace.ID,
+			&trace.ExecutionID,
+			&trace.NodeID,
+			&trace.ActivityName,
+			&trace.Parameters,
+			&trace.Result,
+			&trace.StartedAt,
+			&trace.CompletedAt,
+			&trace.DurationMs,
+			&trace.Attempt,
+			&trace.RetryReason,
+			&trace.Status,
+			&trace.ErrorMessage,
+		)
+		if err != nil {
+			return nil, fmt.Errorf("failed to scan trace: %w", err)
+		}
+		traces = append(traces, trace)
+	}
+
+	return traces, rows.Err()
+}
+
+// GetWorkflowRelations retrieves all relations for a workflow version
+func (db *DB) GetWorkflowRelations(ctx context.Context, workflowID string, version int) ([]WorkflowRelation, error) {
+	var relations []WorkflowRelation
+
+	query := `
+		SELECT id, workflow_id, version, source_node_id, target_node_id, 
+		       relation_type, label, relation_wording, metadata, created_at
+		FROM workflow_relations
+		WHERE workflow_id = $1 AND version = $2
+		ORDER BY created_at DESC
+	`
+
+	rows, err := db.conn.QueryContext(ctx, query, workflowID, version)
+	if err != nil {
+		return nil, fmt.Errorf("failed to query relations: %w", err)
+	}
+	defer rows.Close()
+
+	for rows.Next() {
+		var rel WorkflowRelation
+		if err := rows.Scan(
+			&rel.ID,
+			&rel.WorkflowID,
+			&rel.Version,
+			&rel.SourceNodeID,
+			&rel.TargetNodeID,
+			&rel.RelationType,
+			&rel.Label,
+			&rel.RelationWording,
+			&rel.Metadata,
+			&rel.CreatedAt,
+		); err != nil {
+			return nil, fmt.Errorf("failed to scan relation: %w", err)
+		}
+		relations = append(relations, rel)
+	}
+
+	return relations, rows.Err()
+}
+
+// GetRelationVersions retrieves version history for a specific relation
+func (db *DB) GetRelationVersions(ctx context.Context, workflowID string, edgeID string) ([]WorkflowRelationVersion, error) {
+	var versions []WorkflowRelationVersion
+
+	query := `
+		SELECT id, workflow_id, edge_id, version_num, operation, snapshot, 
+		       changed_at, changed_by, fields_changed
+		FROM workflow_relation_versions
+		WHERE workflow_id = $1 AND edge_id = $2
+		ORDER BY version_num ASC
+	`
+
+	rows, err := db.conn.QueryContext(ctx, query, workflowID, edgeID)
+	if err != nil {
+		return nil, fmt.Errorf("failed to query relation versions: %w", err)
+	}
+	defer rows.Close()
+
+	for rows.Next() {
+		var v WorkflowRelationVersion
+		if err := rows.Scan(
+			&v.ID,
+			&v.WorkflowID,
+			&v.EdgeID,
+			&v.VersionNum,
+			&v.Operation,
+			&v.Snapshot,
+			&v.ChangedAt,
+			&v.ChangedBy,
+			&v.FieldsChanged,
+		); err != nil {
+			return nil, fmt.Errorf("failed to scan version: %w", err)
+		}
+		versions = append(versions, v)
+	}
+
+	return versions, rows.Err()
+}

pkg/db/models.go

@@ -0,0 +1,141 @@
+package db
+
+import (
+	"time"
+)
+
+// WorkflowNode represents a React Flow node in the canvas
+type WorkflowNode struct {
+	ID       string                 `json:"id"`
+	Label    string                 `json:"label"`
+	Type     string                 `json:"type"` // "activity"
+	Position map[string]interface{} `json:"position"`
+	Data     map[string]interface{} `json:"data"`
+}
+
+// WorkflowEdge represents a React Flow edge in the canvas
+type WorkflowEdge struct {
+	ID     string                 `json:"id"`
+	Source string                 `json:"source"`
+	Target string                 `json:"target"`
+	Data   map[string]interface{} `json:"data"`
+}
+
+// Canvas represents the full React Flow canvas (nodes + edges)
+type Canvas struct {
+	Name  string         `json:"name,omitempty"`
+	Nodes []WorkflowNode `json:"nodes"`
+	Edges []WorkflowEdge `json:"edges"`
+}
+
+// Workflow represents a workflow definition in the database
+type Workflow struct {
+	ID            string    `db:"id"`
+	CustomerID    string    `db:"customer_id"`
+	Name          string    `db:"name"`
+	Description   string    `db:"description"`
+	Status        string    `db:"status"`        // "draft", "active", "archived"
+	Version       int       `db:"version"`
+	Nodes         []byte    `db:"nodes"`         // JSONB stored as []byte
+	Edges         []byte    `db:"edges"`         // JSONB stored as []byte
+	CreatedBy     string    `db:"created_by"`
+	CreatedAt     time.Time `db:"created_at"`
+	UpdatedAt     time.Time `db:"updated_at"`
+	LastExecutedAt *time.Time `db:"last_executed_at"`
+}
+
+// WorkflowExecution represents a workflow execution run
+type WorkflowExecution struct {
+	ID           string    `db:"id"`
+	WorkflowID   string    `db:"workflow_id"`
+	CustomerID   string    `db:"customer_id"`
+	TemporalID   string    `db:"temporal_id"` // Temporal execution ID
+	Status       string    `db:"status"`      // "pending", "running", "success", "failed", "cancelled"
+	Inputs       []byte    `db:"inputs"`      // JSONB
+	Outputs      []byte    `db:"outputs"`     // JSONB
+	StartedAt    time.Time `db:"started_at"`
+	CompletedAt  *time.Time `db:"completed_at"`
+	DurationMs   *int      `db:"duration_ms"`
+	ErrorMessage string    `db:"error_message"`
+	ErrorCount   int       `db:"error_count"`
+}
+
+// ExecutionLog represents a detailed activity log entry
+type ExecutionLog struct {
+	ID           int64     `db:"id"`
+	ExecutionID  string    `db:"execution_id"`
+	NodeID       string    `db:"node_id"`      // From canvas node ID
+	ActivityName string    `db:"activity_name"` // "CloneRepo", "AnalyzeCode", etc
+	Level        string    `db:"level"`        // "info", "warn", "error", "debug"
+	Message      string    `db:"message"`
+	Metadata     []byte    `db:"metadata"`    // JSONB
+	LoggedAt     time.Time `db:"logged_at"`
+}
+
+// ActivityTrace represents per-activity execution metrics
+type ActivityTrace struct {
+	ID           int64      `db:"id"`
+	ExecutionID  string     `db:"execution_id"`
+	NodeID       string     `db:"node_id"`
+	ActivityName string     `db:"activity_name"`
+	Parameters   []byte     `db:"parameters"` // JSONB
+	Result       []byte     `db:"result"`     // JSONB
+	StartedAt    time.Time  `db:"started_at"`
+	CompletedAt  *time.Time `db:"completed_at"`
+	DurationMs   *int       `db:"duration_ms"`
+	Attempt      int        `db:"attempt"`
+	RetryReason  string     `db:"retry_reason"`
+	Status       string     `db:"status"` // "running", "success", "failed", "skipped"
+	ErrorMessage string     `db:"error_message"`
+}
+
+// WorkflowStats represents aggregated workflow metrics
+type WorkflowStats struct {
+	WorkflowID        string    `db:"workflow_id"`
+	CustomerID        string    `db:"customer_id"`
+	TotalRuns         int       `db:"total_runs"`
+	SuccessfulRuns    int       `db:"successful_runs"`
+	FailedRuns        int       `db:"failed_runs"`
+	AvgDurationMs     float64   `db:"avg_duration_ms"`
+	MinDurationMs     *int      `db:"min_duration_ms"`
+	MaxDurationMs     *int      `db:"max_duration_ms"`
+	Last30dRuns       int       `db:"last_30d_runs"`
+	Last30dSuccessRate float64  `db:"last_30d_success_rate"`
+	UpdatedAt         time.Time `db:"updated_at"`
+}
+
+// WorkflowMemoryLink represents a connection between execution and memory nodes
+type WorkflowMemoryLink struct {
+	ExecutionID   string    `db:"execution_id"`
+	MemoryNodeSha string    `db:"memory_node_sha"`
+	Relationship  string    `db:"relationship"` // "generated", "used", "learned", "failed_on"
+	CreatedAt     time.Time `db:"created_at"`
+	Notes         string    `db:"notes"`
+}
+
+// WorkflowRelation represents a semantic relation between two canvas nodes
+type WorkflowRelation struct {
+	ID               string    `db:"id"`
+	WorkflowID       string    `db:"workflow_id"`
+	Version          int       `db:"version"`
+	SourceNodeID     string    `db:"source_node_id"`
+	TargetNodeID     string    `db:"target_node_id"`
+	RelationType     string    `db:"relation_type"` // "data-flow", "dependency", "conditional"
+	Label            string    `db:"label"`         // Human-readable relation description
+	RelationWording  []byte    `db:"relation_wording"`  // JSONB with verb, outputs, inputs, confidence
+	Metadata         []byte    `db:"metadata"`      // JSONB for extensibility
+	CreatedAt        time.Time `db:"created_at"`
+}
+
+// WorkflowRelationVersion represents versioned history of relation changes
+type WorkflowRelationVersion struct {
+	ID            string    `db:"id"`
+	WorkflowID    string    `db:"workflow_id"`
+	EdgeID        string    `db:"edge_id"`
+	VersionNum    int       `db:"version_num"`
+	Operation     string    `db:"operation"` // "CREATE", "UPDATE", "DELETE"
+	Snapshot      []byte    `db:"snapshot"` // JSONB full state at this version
+	ChangedAt     time.Time `db:"changed_at"`
+	ChangedBy     string    `db:"changed_by"`
+	FieldsChanged []byte    `db:"fields_changed"` // JSONB array of changed field names
+}

pkg/types/types.go

@@ -0,0 +1,223 @@
+// Package types defines the shared domain model for Poimen workflows.
+// Both workflow/ (orchestration) and activity/ (execution) import from here.
+package types
+
+import (
+	"time"
+
+	"github.com/rockliang/poimen/workflows/pkg/db"
+)
+
+// ===== LLM Configuration =====
+
+type ModelSpec struct {
+	ModelID  string
+	Thinking string // "adaptive" or ""
+	Effort   string // "low", "medium", "high", "xhigh", "max"
+}
+
+type PromptSpec struct {
+	TemplateRef string
+	RawTemplate string
+	Variables   map[string]any
+	Model       ModelSpec
+	LessonsRef  string
+}
+
+type SkillRef struct {
+	Name string
+	URL  string
+}
+
+// ===== Retry & Tuning =====
+
+type PiRetryPolicy struct {
+	ScheduleToCloseTimeout time.Duration
+	InitialInterval        time.Duration
+	MaximumInterval        time.Duration
+	BackoffCoefficient     float64
+	StreamTimeout          time.Duration
+	StreamTimeoutMax       time.Duration
+}
+
+type ActivityTuning struct {
+	ImplementerBaseTimeout  time.Duration
+	ImplementerMaxRetries   int
+	JudgeTimeout            time.Duration
+	PiRetry                 PiRetryPolicy
+	InitialRetryInterval    time.Duration
+	MaxRetryInterval        time.Duration
+	RetryBackoffCoefficient float64
+}
+
+// ===== Orchestrator =====
+
+type OrchestratorConfig struct {
+	SystemPrompt string
+	Skills       []SkillRef
+	RolePrompts  map[string]PromptSpec
+	Tuning       ActivityTuning
+}
+
+type OrchestratorInput struct {
+	TargetRepoPath     string
+	RemoteURL          string
+	Milestone          string
+	Config             OrchestratorConfig
+	DryRun             bool
+	CycleCount         int
+	MaxCyclesBeforeCAN int
+	PiProvider         string
+}
+
+type OrchestratorOutput struct {
+	MilestoneComplete bool
+	Done              bool
+	LastError         string
+}
+
+// ===== TaskUnit =====
+
+type TaskUnitInput struct {
+	TaskID         string
+	RemoteURL      string
+	TargetRepoPath string
+	Milestone      string
+	Config         OrchestratorConfig
+	DryRun         bool
+}
+
+type TaskUnitOutput struct {
+	TaskID   string
+	Status   string
+	Verdict  string
+	Critique string
+	Branch   string
+	Reason   string
+	Changes  string
+}
+
+// ===== Canvas & Relations =====
+
+type RelationWording struct {
+	Verb              string  `json:"verb"`
+	SourceOutput      string  `json:"source_output"`
+	TargetInput       string  `json:"target_input"`
+	ConnectionType    string  `json:"connection_type"`
+	Confidence        float64 `json:"confidence"`
+	SemanticMatch     string  `json:"semantic_match"`
+	TransformerNeeded string  `json:"transformer_needed,omitempty"`
+}
+
+type EdgeWithWording struct {
+	ID              string          `json:"id,omitempty"`
+	Source          string          `json:"source"`
+	Target          string          `json:"target"`
+	RelationType    string          `json:"relation_type"`
+	RelationLabel   string          `json:"relation_label"`
+	RelationWording RelationWording `json:"relation_wording"`
+	CreatedAt       string          `json:"created_at,omitempty"`
+}
+
+type CanvasWithRelationsData struct {
+	WorkflowID string            `json:"workflow_id"`
+	Version    int               `json:"version"`
+	Nodes      []db.WorkflowNode `json:"nodes"`
+	Edges      []db.WorkflowEdge `json:"edges"`
+	Relations  []EdgeWithWording `json:"relations"`
+	UpdatedAt  string            `json:"updated_at"`
+}
+
+// ===== Activity I/O =====
+
+type FetchCanvasRelationsInput struct {
+	WorkflowID string `json:"workflow_id"`
+	Version    int    `json:"version"`
+}
+
+type CanvasReasonerInput struct {
+	Nodes            []db.WorkflowNode `json:"nodes"`
+	Edges            []db.WorkflowEdge `json:"edges"`
+	PreserveExisting bool              `json:"preserve_existing,omitempty"`
+	AuthToken        string            `json:"auth_token,omitempty"`
+}
+
+type QueryPathData struct {
+	SourceID   string   `json:"source_id"`
+	TargetID   string   `json:"target_id"`
+	Distance   int      `json:"distance"`
+	PathCount  int      `json:"path_count"`
+	NodeIDs    []string `json:"node_ids"`
+	Confidence float64  `json:"total_confidence"`
+}
+
+type GraphRAGQueryInput struct {
+	WorkflowID      string                 `json:"workflow_id"`
+	Query           string                 `json:"query"`
+	SearchType      string                 `json:"search_type"`
+	RelationType    string                 `json:"relation_type"`
+	ConfidenceFloor float64                `json:"confidence_floor"`
+	TopK            int                    `json:"top_k"`
+	RankingProfile  string                 `json:"ranking_profile"`
+	Canvas          CanvasWithRelationsData `json:"canvas"`
+}
+
+type GraphRAGQueryOutput struct {
+	WorkflowID  string            `json:"workflow_id"`
+	Query       string            `json:"query"`
+	Edges       []EdgeWithWording `json:"edges"`
+	Paths       []QueryPathData   `json:"paths"`
+	TotalCount  int               `json:"total_count"`
+	HasMore     bool              `json:"has_more"`
+	ExecutionMs int64             `json:"execution_time_ms"`
+}
+
+type CanvasCompatibilityInput struct {
+	Nodes []db.WorkflowNode `json:"nodes"`
+	Edges []db.WorkflowEdge `json:"edges"`
+	Query string            `json:"query,omitempty"`
+}
+
+type IndexGraphRAGInput struct {
+	WorkflowID string            `json:"workflow_id"`
+	Version    int               `json:"version"`
+	Nodes      []db.WorkflowNode `json:"nodes"`
+	Relations  []EdgeWithWording `json:"relations"`
+}
+
+type IndexGraphRAGOutput struct {
+	WorkflowID       string `json:"workflow_id"`
+	Version          int    `json:"version"`
+	IndexedEntities  int    `json:"indexed_entities"`
+	IndexedEdges     int    `json:"indexed_edges"`
+	Status           string `json:"status"`
+	GraphRAGChecksum string `json:"graph_rag_checksum"`
+	IndexedAt        string `json:"indexed_at"`
+}
+
+type PromptUpdate struct {
+	Role string
+	Spec PromptSpec
+}
+
+// ===== Defaults =====
+
+func NewPiRetryPolicy() PiRetryPolicy {
+	return PiRetryPolicy{
+		ScheduleToCloseTimeout: 5 * time.Minute,
+		InitialInterval:        2 * time.Second,
+		MaximumInterval:        30 * time.Second,
+		BackoffCoefficient:     2.0,
+		StreamTimeout:          30 * time.Second,
+		StreamTimeoutMax:       2 * time.Minute,
+	}
+}
+
+func NewActivityTuning() ActivityTuning {
+	return ActivityTuning{
+		ImplementerBaseTimeout: 10 * time.Minute,
+		ImplementerMaxRetries:  3,
+		JudgeTimeout:           5 * time.Minute,
+		PiRetry:                NewPiRetryPolicy(),
+	}
+}

statemachine/types.go

@@ -1,136 +0,0 @@
-package statemachine
-
-import "time"
-
-// ModelSpec defines LLM model configuration.
-type ModelSpec struct {
-	ModelID  string // e.g. "claude-opus-5", "claude-sonnet-5"
-	Thinking string // "adaptive" or ""
-	Effort   string // "low", "medium", "high", "xhigh", "max"
-}
-
-// PromptSpec defines a prompt template with variables and model.
-type PromptSpec struct {
-	TemplateRef string            // e.g. "planner/default.tmpl"
-	RawTemplate string            // overrides TemplateRef if non-empty
-	Variables   map[string]any    // template variables
-	Model       ModelSpec         // which LLM to use
-	LessonsRef  string            // key into lessons store
-}
-
-// PiRetryPolicy defines retry and timeout settings for Pi command execution.
-type PiRetryPolicy struct {
-	ScheduleToCloseTimeout time.Duration // default: 5m
-	InitialInterval        time.Duration // default: 2s
-	MaximumInterval        time.Duration // default: 30s
-	BackoffCoefficient     float64       // default: 2.0
-	StreamTimeout          time.Duration // default: 30s
-	StreamTimeoutMax       time.Duration // default: 2m
-}
-
-// ActivityTuning defines timeouts and retry counts for activities.
-type ActivityTuning struct {
-	ImplementerBaseTimeout time.Duration // default: 10m
-	ImplementerMaxRetries  int           // default: 3
-	JudgeTimeout           time.Duration // default: 5m
-	PiRetry                PiRetryPolicy
-	// Retry policy settings
-	InitialRetryInterval   time.Duration // default: 2s
-	MaxRetryInterval       time.Duration // default: 5m
-	RetryBackoffCoefficient float64      // default: 2.0
-}
-
-// OrchestratorConfig holds all runtime configuration for the orchestrator.
-type OrchestratorConfig struct {
-	SystemPrompt string                    // shared prompt prefix
-	Skills       []SkillRef                // required skill sources
-	RolePrompts  map[string]PromptSpec     // per-role: "planner", "judge", "implementer"
-	Tuning       ActivityTuning
-}
-
-// OrchestratorInput is the input to the Orchestrator workflow.
-type OrchestratorInput struct {
-	TargetRepoPath      string
-	RemoteURL           string
-	Milestone           string // e.g. "T0"
-	Config              OrchestratorConfig
-	DryRun              bool
-	CycleCount          int
-	MaxCyclesBeforeCAN  int // default: 100
-	PiProvider          string // pi provider name (e.g., "local-llm"); required for skill preparation
-}
-
-// OrchestratorOutput is the output of the Orchestrator workflow.
-type OrchestratorOutput struct {
-	MilestoneComplete bool
-	Done              bool
-	LastError         string
-}
-
-// TaskUnitInput is the input to the TaskUnit workflow.
-type TaskUnitInput struct {
-	TaskID         string
-	RemoteURL      string
-	TargetRepoPath string
-	Milestone      string
-	Config         OrchestratorConfig
-	DryRun         bool
-}
-
-// TaskUnitOutput is the output of the TaskUnit workflow.
-type TaskUnitOutput struct {
-	TaskID   string
-	Status   string // "success" or "failed"
-	Verdict  string // "pass" or "fail" from judge
-	Critique string // feedback from judge
-	Branch   string
-	Reason   string // error reason if failed
-	Changes  string // summary of changes
-}
-
-// SkillRef references a skill source.
-type SkillRef struct {
-	Name string // skill identifier
-	URL  string // source to clone
-}
-
-// Default values for types.
-const (
-	defaultScheduleToCloseTimeout = 5 * time.Minute
-	defaultInitialInterval         = 2 * time.Second
-	defaultMaximumInterval         = 30 * time.Second
-	defaultBackoffCoefficient      = 2.0
-	defaultStreamTimeout           = 30 * time.Second
-	defaultStreamTimeoutMax        = 2 * time.Minute
-	defaultImplementerBaseTimeout  = 10 * time.Minute
-	defaultImplementerMaxRetries   = 3
-	defaultJudgeTimeout            = 5 * time.Minute
-)
-
-// NewPiRetryPolicy returns a PiRetryPolicy with defaults.
-func NewPiRetryPolicy() PiRetryPolicy {
-	return PiRetryPolicy{
-		ScheduleToCloseTimeout: defaultScheduleToCloseTimeout,
-		InitialInterval:        defaultInitialInterval,
-		MaximumInterval:        defaultMaximumInterval,
-		BackoffCoefficient:     defaultBackoffCoefficient,
-		StreamTimeout:          defaultStreamTimeout,
-		StreamTimeoutMax:       defaultStreamTimeoutMax,
-	}
-}
-
-// NewActivityTuning returns an ActivityTuning with defaults.
-func NewActivityTuning() ActivityTuning {
-	return ActivityTuning{
-		ImplementerBaseTimeout: defaultImplementerBaseTimeout,
-		ImplementerMaxRetries:  defaultImplementerMaxRetries,
-		JudgeTimeout:           defaultJudgeTimeout,
-		PiRetry:                NewPiRetryPolicy(),
-	}
-}
-
-// PromptUpdate represents an update to a role prompt.
-type PromptUpdate struct {
-	Role string
-	Spec PromptSpec
-}

tests/git_test.go

@@ -9,7 +9,7 @@ import (
 	"testing"
 
 	"github.com/stretchr/testify/assert"
-	"github.com/rockliang/poimen/workflows/action"
+	"github.com/rockliang/poimen/workflows/activity"
 )
 
 func TestGitCloneAndFetch(t *testing.T) {
@@ -56,7 +56,7 @@ func TestGitCloneAndFetch(t *testing.T) {
 
 	// Test clone into empty path
 	ctx := context.Background()
-	err := action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err := activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: targetDir,
 	})
@@ -89,7 +89,7 @@ func TestGitCloneAndFetch(t *testing.T) {
 	}
 
 	// Test fetch on existing repo
-	err = action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err = activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: targetDir,
 	})
@@ -139,14 +139,14 @@ func TestGitWorktreeAdd(t *testing.T) {
 
 	// Clone the repo
 	ctx := context.Background()
-	err := action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err := activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: repoDir,
 	})
 	assert.NoError(t, err, "clone should succeed")
 
 	// Test worktree add
-	worktreePath, err := action.GitWorktreeAddActivity(ctx, action.GitWorktreeAddInput{
+	worktreePath, err := activity.GitWorktreeAddActivity(ctx, activity.GitWorktreeAddInput{
 		RepoPath: repoDir,
 		TaskID:   "T0.1",
 	})
@@ -207,14 +207,14 @@ func TestGitCommit(t *testing.T) {
 
 	// Clone the repo
 	ctx := context.Background()
-	err := action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err := activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: repoDir,
 	})
 	assert.NoError(t, err, "clone should succeed")
 
 	// Create a worktree
-	worktreePath, err := action.GitWorktreeAddActivity(ctx, action.GitWorktreeAddInput{
+	worktreePath, err := activity.GitWorktreeAddActivity(ctx, activity.GitWorktreeAddInput{
 		RepoPath: repoDir,
 		TaskID:   "T0.1",
 	})
@@ -227,7 +227,7 @@ func TestGitCommit(t *testing.T) {
 	}
 
 	// Commit changes
-	err = action.GitCommitActivity(ctx, action.GitCommitInput{
+	err = activity.GitCommitActivity(ctx, activity.GitCommitInput{
 		WorktreePath: worktreePath,
 		Message:      "Add new file",
 	})
@@ -283,14 +283,14 @@ func TestGitDiff(t *testing.T) {
 
 	// Clone the repo
 	ctx := context.Background()
-	err := action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err := activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: repoDir,
 	})
 	assert.NoError(t, err, "clone should succeed")
 
 	// Create a worktree
-	worktreePath, err := action.GitWorktreeAddActivity(ctx, action.GitWorktreeAddInput{
+	worktreePath, err := activity.GitWorktreeAddActivity(ctx, activity.GitWorktreeAddInput{
 		RepoPath: repoDir,
 		TaskID:   "T0.1",
 	})
@@ -309,7 +309,7 @@ func TestGitDiff(t *testing.T) {
 	}
 
 	// Get diff (should show the staged change)
-	diffOutput, err := action.GitDiffActivity(ctx, action.GitDiffInput{
+	diffOutput, err := activity.GitDiffActivity(ctx, activity.GitDiffInput{
 		WorktreePath: worktreePath,
 	})
 	assert.NoError(t, err, "diff should succeed")
@@ -378,7 +378,7 @@ func TestGitSquashMerge(t *testing.T) {
 
 	// Clone for the orchestrator to use
 	ctx := context.Background()
-	err := action.CloneRepoActivity(ctx, action.CloneRepoInput{
+	err := activity.CloneRepoActivity(ctx, activity.CloneRepoInput{
 		RemoteURL:      sourceDir,
 		TargetRepoPath: repoDir,
 	})
@@ -387,7 +387,7 @@ func TestGitSquashMerge(t *testing.T) {
 	// Create multiple worktrees with changes
 	for i := 1; i <= 2; i++ {
 		taskID := fmt.Sprintf("T0.%d", i)
-		worktreePath, err := action.GitWorktreeAddActivity(ctx, action.GitWorktreeAddInput{
+		worktreePath, err := activity.GitWorktreeAddActivity(ctx, activity.GitWorktreeAddInput{
 			RepoPath: repoDir,
 			TaskID:   taskID,
 		})
@@ -400,7 +400,7 @@ func TestGitSquashMerge(t *testing.T) {
 		}
 
 		// Commit changes
-		err = action.GitCommitActivity(ctx, action.GitCommitInput{
+		err = activity.GitCommitActivity(ctx, activity.GitCommitInput{
 			WorktreePath: worktreePath,
 			Message:      fmt.Sprintf("Task %s implementation", taskID),
 		})
@@ -408,7 +408,7 @@ func TestGitSquashMerge(t *testing.T) {
 	}
 
 	// Perform squash merge
-	err = action.GitSquashMergeActivity(ctx, action.GitSquashMergeInput{
+	err = activity.GitSquashMergeActivity(ctx, activity.GitSquashMergeInput{
 		RepoPath: repoDir,
 		Branches: []string{"task/T0.1", "task/T0.2"},
 		Message:  "Milestone T0: completed all tasks",

tests/routing_workflow_test.go

@@ -6,7 +6,7 @@ import (
 	"testing"
 
 	"github.com/rockliang/poimen/workflows/internal/routing"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 	"github.com/stretchr/testify/require"
 	"go.temporal.io/sdk/testsuite"
 )
@@ -45,14 +45,14 @@ func TestRoutingWorkflow_SimpleWorkflow(t *testing.T) {
 		},
 	}
 
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	require.Equal(t, "COMPLETED", output.Status)
 	require.NotNil(t, output.FinalOutput)
@@ -96,14 +96,14 @@ func TestRoutingWorkflow_MultiStepWorkflow(t *testing.T) {
 		},
 	}
 
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	t.Logf("Output: %+v", output)
 	t.Logf("Error: %s", output.Error)
@@ -130,14 +130,14 @@ func TestRoutingWorkflow_PassState(t *testing.T) {
 		},
 	}
 
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	require.Equal(t, "COMPLETED", output.Status)
 }
@@ -160,14 +160,14 @@ func TestRoutingWorkflow_FailState(t *testing.T) {
 		},
 	}
 
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	require.Equal(t, "FAILED", output.Status)
 	require.Contains(t, output.Error, "WorkflowError")
@@ -219,14 +219,14 @@ func TestRoutingWorkflow_ErrorCatch(t *testing.T) {
 		},
 	}
 
-	input := statemachine.RoutingWorkflowInput{Spec: spec}
+	input := workflow.RoutingWorkflowInput{Spec: spec}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	require.Equal(t, "FAILED", output.Status)
 	require.Contains(t, output.Error, "CaughtError")
@@ -237,14 +237,14 @@ func TestRoutingWorkflow_EmptySpec(t *testing.T) {
 	env := testSuite.NewTestWorkflowEnvironment()
 
 	// Empty spec
-	input := statemachine.RoutingWorkflowInput{Spec: nil}
+	input := workflow.RoutingWorkflowInput{Spec: nil}
 
-	env.ExecuteWorkflow(statemachine.RoutingWorkflow, input)
+	env.ExecuteWorkflow(workflow.RoutingWorkflow, input)
 
 	require.True(t, env.IsWorkflowCompleted())
 	require.NoError(t, env.GetWorkflowError())
 
-	var output statemachine.RoutingWorkflowOutput
+	var output workflow.RoutingWorkflowOutput
 	require.NoError(t, env.GetWorkflowResult(&output))
 	require.Equal(t, "FAILED", output.Status)
 	require.Contains(t, output.Error, "empty")

tests/temporal_integration_test.go

@@ -9,7 +9,7 @@ import (
 	"github.com/stretchr/testify/assert"
 	"go.temporal.io/sdk/client"
 	"github.com/rockliang/poimen/workflows/internal/config"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 )
 
 // TestTemporalConnection verifies the worker is connected and healthy
@@ -67,7 +67,7 @@ func TestActivityExecution(t *testing.T) {
 	runResp, err := c.ExecuteWorkflow(ctx, client.StartWorkflowOptions{
 		ID:        workflowID,
 		TaskQueue: "poimen-taskqueue",
-	}, statemachine.TestWorkflow)
+	}, workflow.TestWorkflow)
 
 	assert.NoError(t, err, "failed to execute test workflow")
 	assert.NotNil(t, runResp, "workflow response should not be nil")
@@ -136,28 +136,28 @@ func TestOrchestratorWorkflowIntegration(t *testing.T) {
 	defer cancel()
 
 	// Create minimal orchestrator input
-	input := statemachine.OrchestratorInput{
+	input := workflow.OrchestratorInput{
 		RemoteURL:      "https://forgejo.riotpiao.com/rock/poimen",
 		TargetRepoPath: "/tmp/test-poimen-integration",
 		Milestone:      "T0",
-		Config: statemachine.OrchestratorConfig{
+		Config: workflow.OrchestratorConfig{
 			SystemPrompt: "You are a code generation assistant. Generate simple test code.",
-			RolePrompts: map[string]statemachine.PromptSpec{
+			RolePrompts: map[string]workflow.PromptSpec{
 				"planner": {
 					TemplateRef: "planner/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID: "ornith",
 					},
 				},
 				"judge": {
 					TemplateRef: "judge/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID: "ornith",
 					},
 				},
 				"implementer": {
 					TemplateRef: "implementer/default.tmpl",
-					Model: statemachine.ModelSpec{
+					Model: workflow.ModelSpec{
 						ModelID: "claude-sonnet-5",
 					},
 				},
@@ -170,7 +170,7 @@ func TestOrchestratorWorkflowIntegration(t *testing.T) {
 	runResp, err := c.ExecuteWorkflow(ctx, client.StartWorkflowOptions{
 		ID:        workflowID,
 		TaskQueue: "poimen-taskqueue",
-	}, statemachine.OrchestratorWorkflow, input)
+	}, workflow.OrchestratorWorkflow, input)
 
 	assert.NoError(t, err, "failed to execute orchestrator workflow")
 	t.Logf("✅ Orchestrator workflow started: %s", workflowID)

tests/temporal_routing_test.go

@@ -9,7 +9,7 @@ import (
 	"time"
 
 	"github.com/rockliang/poimen/workflows/internal/routing"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 	"github.com/stretchr/testify/require"
 	"go.temporal.io/sdk/client"
 )
@@ -69,12 +69,12 @@ func TestTemporalRoutingWorkflow(t *testing.T) {
 
 		// Submit to Temporal
 		workflowID := "test-routing-" + time.Now().Format("20060102-150405")
-		input := statemachine.RoutingWorkflowInput{Spec: output.Spec}
+		input := workflow.RoutingWorkflowInput{Spec: output.Spec}
 
 		run, err := c.ExecuteWorkflow(ctx, client.StartWorkflowOptions{
 			ID:        workflowID,
 			TaskQueue: "poimen-taskqueue",
-		}, statemachine.RoutingWorkflow, input)
+		}, workflow.RoutingWorkflow, input)
 		require.NoError(t, err)
 
 		t.Logf("Workflow submitted: ID=%s, RunID=%s", run.GetID(), run.GetRunID())
@@ -118,18 +118,18 @@ func TestTemporalRoutingWorkflow(t *testing.T) {
 		}
 
 		workflowID := "test-pass-only-" + time.Now().Format("20060102-150405")
-		input := statemachine.RoutingWorkflowInput{Spec: spec}
+		input := workflow.RoutingWorkflowInput{Spec: spec}
 
 		run, err := c.ExecuteWorkflow(ctx, client.StartWorkflowOptions{
 			ID:        workflowID,
 			TaskQueue: "poimen-taskqueue",
-		}, statemachine.RoutingWorkflow, input)
+		}, workflow.RoutingWorkflow, input)
 		require.NoError(t, err)
 
 		t.Logf("Pass-only workflow submitted: ID=%s", run.GetID())
 
 		// Wait for result (Pass states don't need workers)
-		var result statemachine.RoutingWorkflowOutput
+		var result workflow.RoutingWorkflowOutput
 		err = run.Get(ctx, &result)
 		require.NoError(t, err)
 

tests/types_test.go

@@ -5,12 +5,12 @@ import (
 	"time"
 
 	"github.com/stretchr/testify/assert"
-	"github.com/rockliang/poimen/workflows/statemachine"
+	"github.com/rockliang/poimen/workflows/workflow"
 )
 
 func TestTypesDefaults(t *testing.T) {
 	// Test PiRetryPolicy defaults
-	pr := statemachine.NewPiRetryPolicy()
+	pr := workflow.NewPiRetryPolicy()
 	assert.Equal(t, 5*time.Minute, pr.ScheduleToCloseTimeout, "ScheduleToCloseTimeout should be 5m")
 	assert.Equal(t, 2*time.Second, pr.InitialInterval, "InitialInterval should be 2s")
 	assert.Equal(t, 30*time.Second, pr.MaximumInterval, "MaximumInterval should be 30s")
@@ -19,7 +19,7 @@ func TestTypesDefaults(t *testing.T) {
 	assert.Equal(t, 2*time.Minute, pr.StreamTimeoutMax, "StreamTimeoutMax should be 2m")
 
 	// Test ActivityTuning defaults
-	at := statemachine.NewActivityTuning()
+	at := workflow.NewActivityTuning()
 	assert.Equal(t, 10*time.Minute, at.ImplementerBaseTimeout, "ImplementerBaseTimeout should be 10m")
 	assert.Equal(t, 3, at.ImplementerMaxRetries, "ImplementerMaxRetries should be 3")
 	assert.Equal(t, 5*time.Minute, at.JudgeTimeout, "JudgeTimeout should be 5m")
@@ -31,7 +31,7 @@ func TestTypesDefaults(t *testing.T) {
 }
 
 func TestModelSpec(t *testing.T) {
-	spec := statemachine.ModelSpec{
+	spec := workflow.ModelSpec{
 		ModelID:  "claude-opus-5",
 		Thinking: "adaptive",
 		Effort:   "high",
@@ -42,13 +42,13 @@ func TestModelSpec(t *testing.T) {
 }
 
 func TestPromptSpec(t *testing.T) {
-	spec := statemachine.PromptSpec{
+	spec := workflow.PromptSpec{
 		TemplateRef: "planner/default.tmpl",
 		RawTemplate: "",
 		Variables: map[string]any{
 			"key": "value",
 		},
-		Model: statemachine.ModelSpec{
+		Model: workflow.ModelSpec{
 			ModelID: "claude-opus-5",
 		},
 		LessonsRef: "T0.1",
@@ -61,18 +61,18 @@ func TestPromptSpec(t *testing.T) {
 }
 
 func TestOrchestratorConfig(t *testing.T) {
-	cfg := statemachine.OrchestratorConfig{
+	cfg := workflow.OrchestratorConfig{
 		SystemPrompt: "You are an expert",
-		Skills: []statemachine.SkillRef{
+		Skills: []workflow.SkillRef{
 			{Name: "golang-skills", URL: "https://example.com/skill1"},
 		},
-		RolePrompts: map[string]statemachine.PromptSpec{
+		RolePrompts: map[string]workflow.PromptSpec{
 			"planner": {
 				TemplateRef: "planner/default.tmpl",
-				Model: statemachine.ModelSpec{ModelID: "claude-opus-5"},
+				Model: workflow.ModelSpec{ModelID: "claude-opus-5"},
 			},
 		},
-		Tuning: statemachine.NewActivityTuning(),
+		Tuning: workflow.NewActivityTuning(),
 	}
 	assert.Equal(t, "You are an expert", cfg.SystemPrompt)
 	assert.Len(t, cfg.Skills, 1)
@@ -81,7 +81,7 @@ func TestOrchestratorConfig(t *testing.T) {
 }
 
 func TestTaskUnitInput(t *testing.T) {
-	input := statemachine.TaskUnitInput{
+	input := workflow.TaskUnitInput{
 		TaskID:         "T0.1",
 		RemoteURL:      "https://github.com/example/repo",
 		TargetRepoPath: "/tmp/repo",

workflow/orchestrator.go

@@ -1,4 +1,4 @@
-package statemachine
+package workflow
 
 import (
 	"fmt"

workflow/orchestrator_recovery.go

@@ -1,4 +1,4 @@
-package statemachine
+package workflow
 
 import (
 	"fmt"

workflow/routing_workflow.go

@@ -1,4 +1,4 @@
-package statemachine
+package workflow
 
 import (
 	"fmt"

workflow/signals.go

@@ -1,3 +1,3 @@
-package statemachine
+package workflow
 
 // Empty stub - will be filled in T0.7

workflow/taskunit.go

@@ -1,4 +1,4 @@
-package statemachine
+package workflow
 
 import (
 	"fmt"

workflow/test_workflow.go

@@ -1,4 +1,4 @@
-package statemachine
+package workflow
 
 import (
 	"go.temporal.io/sdk/workflow"

workflow/types.go

@@ -0,0 +1,20 @@
+package workflow
+
+import "github.com/rockliang/poimen/workflows/pkg/types"
+
+// Re-export from pkg/types — single source of truth.
+type ModelSpec = types.ModelSpec
+type PromptSpec = types.PromptSpec
+type SkillRef = types.SkillRef
+type PiRetryPolicy = types.PiRetryPolicy
+type ActivityTuning = types.ActivityTuning
+type OrchestratorConfig = types.OrchestratorConfig
+type OrchestratorInput = types.OrchestratorInput
+type OrchestratorOutput = types.OrchestratorOutput
+type TaskUnitInput = types.TaskUnitInput
+type TaskUnitOutput = types.TaskUnitOutput
+type PromptUpdate = types.PromptUpdate
+type EdgeWithWording = types.EdgeWithWording
+
+var NewPiRetryPolicy = types.NewPiRetryPolicy
+var NewActivityTuning = types.NewActivityTuning

workflow/workflow_graph_query.go

@@ -0,0 +1,103 @@
+package workflow
+
+import (
+	"time"
+
+	"go.temporal.io/sdk/temporal"
+	"go.temporal.io/sdk/workflow"
+
+	"github.com/rockliang/poimen/workflows/pkg/types"
+)
+
+type WorkflowGraphQueryInput struct {
+	WorkflowID       string  `json:"workflow_id"`
+	Query            string  `json:"query"`
+	SearchType       string  `json:"search_type"`
+	RelationType     string  `json:"relation_type"`
+	Version          int     `json:"version"`
+	ConfidenceFloor  float64 `json:"confidence_floor"`
+	TopK             int     `json:"top_k"`
+	FindPaths        bool    `json:"find_paths"`
+	TargetNodeID     string  `json:"target_node_id"`
+	MaxPathDepth     int     `json:"max_path_depth"`
+	RankingProfile   string  `json:"ranking_profile"`
+	IncludeReasoning bool    `json:"include_reasoning"`
+}
+
+type WorkflowGraphQueryOutput struct {
+	WorkflowID      string                  `json:"workflow_id"`
+	Query           string                  `json:"query"`
+	Version         int                     `json:"version"`
+	ExecutionTimeMs int64                   `json:"execution_time_ms"`
+	Results         []types.EdgeWithWording `json:"results"`
+	Paths           []QueryPath             `json:"paths"`
+	TotalCount      int                     `json:"total_count"`
+	HasMore         bool                    `json:"has_more"`
+	RankingProfile  string                  `json:"ranking_profile"`
+}
+
+type QueryPath struct {
+	SourceID   string   `json:"source_id"`
+	TargetID   string   `json:"target_id"`
+	Distance   int      `json:"distance"`
+	PathCount  int      `json:"path_count"`
+	NodeIDs    []string `json:"node_ids"`
+	Confidence float64  `json:"total_confidence"`
+}
+
+func WorkflowGraphQuery(ctx workflow.Context, input WorkflowGraphQueryInput) (WorkflowGraphQueryOutput, error) {
+	startTime := time.Now()
+	output := WorkflowGraphQueryOutput{
+		WorkflowID:     input.WorkflowID,
+		Query:          input.Query,
+		Version:        input.Version,
+		RankingProfile: input.RankingProfile,
+		Results:        []types.EdgeWithWording{},
+		Paths:          []QueryPath{},
+	}
+
+	opts := workflow.ActivityOptions{
+		StartToCloseTimeout: 120 * time.Second,
+		RetryPolicy: &temporal.RetryPolicy{
+			InitialInterval:    2 * time.Second,
+			BackoffCoefficient: 2.0,
+			MaximumInterval:    10 * time.Second,
+			MaximumAttempts:    3,
+		},
+	}
+	ctx = workflow.WithActivityOptions(ctx, opts)
+
+	var canvasData types.CanvasWithRelationsData
+	err := workflow.ExecuteActivity(ctx, "FetchCanvasRelationsActivity",
+		types.FetchCanvasRelationsInput{
+			WorkflowID: input.WorkflowID,
+			Version:    input.Version,
+		},
+	).Get(ctx, &canvasData)
+	if err != nil {
+		return output, err
+	}
+
+	var graphResults types.GraphRAGQueryOutput
+	err = workflow.ExecuteActivity(ctx, "QueryGraphRAGActivity",
+		types.GraphRAGQueryInput{
+			WorkflowID:      input.WorkflowID,
+			Query:           input.Query,
+			SearchType:      input.SearchType,
+			RelationType:    input.RelationType,
+			ConfidenceFloor: input.ConfidenceFloor,
+			TopK:            input.TopK,
+			RankingProfile:  input.RankingProfile,
+			Canvas:          canvasData,
+		},
+	).Get(ctx, &graphResults)
+	if err != nil {
+		return output, err
+	}
+
+	output.Results = graphResults.Edges
+	output.TotalCount = graphResults.TotalCount
+	output.HasMore = graphResults.HasMore
+	output.ExecutionTimeMs = time.Since(startTime).Milliseconds()
+	return output, nil
+}