Lang graph

.gitignore

@@ -117,4 +117,4 @@ Thumbs.db
 .playwright-mcp/
 .stories/
 .testplans/
-.spikes/
+.spikes/

BACKEND_BUILD_FIX.md

@@ -0,0 +1,65 @@
+# Fix Backend Exec Format Error
+
+## Problem
+`exec container process '/app/./main': Exec format error` - This means the binary architecture doesn't match the cluster.
+
+## Solution: Rebuild with explicit platform
+
+### Option 1: Build with --platform flag (Recommended)
+
+```bash
+cd components/backend
+
+# Build explicitly for linux/amd64
+podman build --platform linux/amd64 -t quay.io/gkrumbach07/vteam_backend:langgraph-mvp .
+
+# Push
+podman push quay.io/gkrumbach07/vteam_backend:langgraph-mvp
+```
+
+### Option 2: Update Dockerfile (Already done)
+
+The Dockerfile has been updated to explicitly build for `GOARCH=amd64`. Just rebuild:
+
+```bash
+cd components/backend
+podman build --platform linux/amd64 -t quay.io/gkrumbach07/vteam_backend:langgraph-mvp .
+podman push quay.io/gkrumbach07/vteam_backend:langgraph-mvp
+```
+
+### Option 3: Check your VM architecture
+
+```bash
+# Check what architecture your VM is
+uname -m
+
+# If it's aarch64 (ARM64), you MUST use --platform linux/amd64
+# If it's x86_64 (AMD64), the build should work, but still use --platform to be safe
+```
+
+## After Rebuilding
+
+```bash
+# Update deployment
+oc set image deployment/backend-api backend-api=quay.io/gkrumbach07/vteam_backend:langgraph-mvp -n ambient-code
+
+# Wait for rollout
+oc rollout status deployment/backend-api -n ambient-code
+
+# Check logs
+oc logs -n ambient-code -l app=backend-api --tail=50
+```
+
+## Verify the Binary Architecture
+
+You can check the binary architecture inside the image:
+
+```bash
+# Create a test container
+podman run --rm quay.io/gkrumbach07/vteam_backend:langgraph-mvp file /app/main
+
+# Should show: ELF 64-bit LSB executable, x86-64
+# If it shows ARM or other architecture, rebuild with --platform linux/amd64
+```
+
+

BUILD_COMMANDS.md

@@ -0,0 +1,128 @@
+# Build and Push Commands
+
+## Prerequisites
+
+```bash
+# Login to Quay.io (do this once)
+podman login quay.io
+# Enter username: gkrumbach07
+# Enter password/token when prompted
+```
+
+---
+
+## 1. LangGraph Wrapper Base Image
+
+**Location:** `components/runners/langgraph-wrapper/`
+
+```bash
+cd components/runners/langgraph-wrapper
+
+# Build for linux/amd64 (required for OpenShift/K8s)
+podman build --platform linux/amd64 -t quay.io/gkrumbach07/langgraph-wrapper:base .
+
+# Push to Quay.io
+podman push quay.io/gkrumbach07/langgraph-wrapper:base
+
+# Get digest (save this for reference)
+podman inspect quay.io/gkrumbach07/langgraph-wrapper:base | jq -r '.[0].RepoDigests[0]'
+```
+
+**Note:** Make sure this image is public in Quay.io settings so workflows can pull it.
+
+---
+
+## 2. Backend Image
+
+**Location:** `components/backend/`
+
+**On your Linux VM:**
+
+```bash
+cd components/backend
+
+# Build the backend image
+podman build -t quay.io/gkrumbach07/vteam_backend:langgraph-mvp .
+
+# Push to Quay.io
+podman push quay.io/gkrumbach07/vteam_backend:langgraph-mvp
+
+# Get digest (save this for reference)
+podman inspect quay.io/gkrumbach07/vteam_backend:langgraph-mvp | jq -r '.[0].RepoDigests[0]'
+```
+
+**After pushing, update the deployment:**
+
+```bash
+# Update deployment to use new image
+oc set image deployment/backend-api backend-api=quay.io/gkrumbach07/vteam_backend:langgraph-mvp -n ambient-code
+
+# Or use digest for pinning:
+oc set image deployment/backend-api backend-api=quay.io/gkrumbach07/vteam_backend@sha256:YOUR_DIGEST -n ambient-code
+```
+
+---
+
+## 3. Example Workflow Image (Optional - for testing)
+
+**Location:** `test-workflow-example/`
+
+```bash
+cd test-workflow-example
+
+# Build for linux/amd64
+podman build --platform linux/amd64 -t quay.io/gkrumbach07/langgraph-example-workflow:v1.0.0 .
+
+# Push to Quay.io
+podman push quay.io/gkrumbach07/langgraph-example-workflow:v1.0.0
+
+# Get digest (required for registration)
+podman inspect quay.io/gkrumbach07/langgraph-example-workflow:v1.0.0 | jq -r '.[0].RepoDigests[0]'
+```
+
+**After pushing, register in the backend:**
+
+```bash
+# Replace YOUR_DIGEST with the digest from above
+curl -X POST "http://localhost:8080/api/projects/YOUR_PROJECT/workflows/example-workflow/versions" \
+  -H "Content-Type: application/json" \
+  -d '{
+    "version": "v1.0.0",
+    "imageRef": "quay.io/gkrumbach07/langgraph-example-workflow@sha256:YOUR_DIGEST",
+    "graphs": [
+      {
+        "name": "main",
+        "entry": "app.workflow:build_app"
+      }
+    ],
+    "inputsSchema": {
+      "type": "object",
+      "properties": {
+        "message": {
+          "type": "string",
+          "description": "Input message to process"
+        }
+      },
+      "required": ["message"]
+    }
+  }'
+```
+
+---
+
+## Quick Reference
+
+### Image Tags:
+- Base wrapper: `quay.io/gkrumbach07/langgraph-wrapper:base`
+- Backend: `quay.io/gkrumbach07/vteam_backend:langgraph-mvp`
+- Example workflow: `quay.io/gkrumbach07/langgraph-example-workflow:v1.0.0`
+
+### Registry Paths:
+- Your personal org: `quay.io/gkrumbach07/*`
+- Main org: `quay.io/ambient_code/*`
+
+### Build Order:
+1. **First:** Build and push `langgraph-wrapper:base` (workflows depend on it)
+2. **Second:** Build and push `vteam_backend:langgraph-mvp` (system component)
+3. **Third:** Build and push example workflow (for testing)
+

DEPLOY_CHECKLIST.md

@@ -0,0 +1,114 @@
+# Deployment Checklist for LangGraph MVP
+
+## ✅ Already Done
+- ✅ Logged into OpenShift cluster (as `clusteradmin`)
+- ✅ Built `langgraph-wrapper:base` image
+- ✅ Built example workflow image
+
+## 🔨 What You Need to Do
+
+### 1. Build All Component Images
+
+The LangGraph MVP requires these images to be built and deployed:
+- **Backend** (includes Postgres DB code)
+- **Frontend** 
+- **Operator** (handles LangGraph workflow pods)
+- **Runner** (legacy Claude runner)
+
+```bash
+# Build all images (using podman, targeting linux/amd64)
+make build-all CONTAINER_ENGINE=podman PLATFORM=linux/amd64
+```
+
+### 2. Push Images to Registry
+
+**Option A: Use Existing Images from quay.io/ambient_code**
+- Skip building/pushing if you trust the existing images
+- The deploy script uses `quay.io/ambient_code/*:latest` by default
+
+**Option B: Build and Push Your Own**
+```bash
+# Set your registry
+export REGISTRY="quay.io/gkrumbach07"
+
+# Tag and push
+make push-all CONTAINER_ENGINE=podman REGISTRY=$REGISTRY
+```
+
+### 3. Deploy to Cluster
+
+```bash
+# From project root
+cd components/manifests
+
+# Copy env file if first time
+cp env.example .env
+
+# Edit .env and set at least:
+# - ANTHROPIC_API_KEY=your-key
+# - (Optional) CONTAINER_REGISTRY if using custom images
+
+# Deploy (from project root)
+make deploy
+```
+
+The deploy script will:
+- ✅ Check you're logged in (already done)
+- ✅ Create namespace `ambient-code`
+- ✅ Deploy Postgres (new!)
+- ✅ Deploy Backend, Frontend, Operator
+- ✅ Set up RBAC, CRDs, Routes
+
+### 4. Verify Deployment
+
+```bash
+# Check pods
+oc get pods -n ambient-code
+
+# Wait for all pods to be Running
+oc get pods -n ambient-code -w
+
+# Check Postgres is running (NEW!)
+oc get pods -n ambient-code | grep postgres
+
+# Check services
+oc get services -n ambient-code
+
+# Get frontend route
+oc get route frontend-route -n ambient-code
+```
+
+## 📝 Important Notes
+
+1. **LangGraph Wrapper**: Already built and pushed to `quay.io/gkrumbach07/langgraph-wrapper:base` ✅
+   - This is the base image users extend for their workflows
+   - Doesn't need to be in Makefile - it's a separate concern
+
+2. **Postgres**: Will be deployed automatically via `make deploy`
+   - New StatefulSet, Service, Secret, ConfigMap
+   - Backend connects to it automatically
+
+3. **Cluster Login**: Already done ✅
+   - `oc whoami` shows `clusteradmin`
+   - `make deploy` will work
+
+4. **Image Registry**: 
+   - Default: Uses `quay.io/ambient_code/*:latest`
+   - Custom: Set `CONTAINER_REGISTRY` in `.env` or override in deploy script
+
+## 🚀 Quick Start
+
+```bash
+# 1. Build images (if needed)
+make build-all CONTAINER_ENGINE=podman PLATFORM=linux/amd64
+
+# 2. Deploy everything
+make deploy
+
+# 3. Watch pods start
+oc get pods -n ambient-code -w
+
+# 4. Get frontend URL
+oc get route frontend-route -n ambient-code
+```
+