Remove all MCP-related infrastructure in favor of direct REST API calls. MCP added layers without value: Docker container, stdio bridge, hybrid fallback, healthchecks, SSE transport — all of which added failure modes and token overhead. Deleted: - docker/mcp-gitea/docker-compose.yml (MCP container config) - scripts/mcp-gitea-stdio.cjs (stdio bridge) - scripts/e2e-mcp-stdio-test*.py (MCP E2E tests) - scripts/test-kilo-mcp-integration.py - src/kilocode/agent-manager/mcp-gitea-client.ts (548 lines of MCP wrapper) - MCP-STDIO-SETUP.md (MCP documentation) - .vscode/settings.json (hardcoded MCP config with token) - .kilo/skills/mcp-gitea-connection/ and mcp-gitea.research.md Restored: - pipeline-runner.ts: HybridGiteaClient → GiteaClient (direct REST) Removed MCP dependency, imports, and initialization. No healthcheck waits, no container startup delays. - process-continuity.md: removed MCP-specific failure modes - e2e-gns2-test.py: removed Basic Auth, use token auth; fixed spec reference
116 lines
3.7 KiB
Markdown
116 lines
3.7 KiB
Markdown
# GNS-2: Process Continuity Rules
|
|
|
|
## Problem
|
|
|
|
The pipeline repeatedly broke in early GNS-2 phases because:
|
|
1. **service_healthy deadlock** (docker-compose.yml) — container couldn't start because it was waiting for its own healthcheck to pass before it was running
|
|
2. **Network overlap** — subnet 172.28.0.0/16 conflicted with existing Docker networks
|
|
3. **Hardcoded IPs and ports** — rigid Docker setups caused conflicts with host networks
|
|
4. **Operator dependency** — process stopped when technical barriers hit, required human decisions
|
|
|
|
## Root Cause
|
|
|
|
| Failure | Why it happened | Operator-Free Fix |
|
|
|---------|-----------------|-----------------|
|
|
| `service_healthy` deadlock | Docker compose blocked startup waiting for healthcheck on a container that wasn't yet running | Use `condition: service_started` for depends_on |
|
|
| Subnet `172.28.0.0/16` conflict | Hardcoded IP overlap with host Docker networks | Remove `ipam` config, let Docker auto-assign |
|
|
| Rigid container configs | Inflexible Docker Compose setups caused conflicts with host networks | Use dynamic networking and auto-assigned IPs |
|
|
| `/health` endpoint mismatch | Container used unstable `/health` endpoint | Probe guaranteed endpoints or use `service_started` |
|
|
|
|
## Operator-Free Design Principles
|
|
|
|
### 1. No `service_healthy` Conditions
|
|
```yaml
|
|
# PROBLEM: deadlock
|
|
depends_on:
|
|
service:
|
|
condition: service_healthy # Container waits for itself
|
|
|
|
# FIX: allow startup, healthcheck as observer only
|
|
depends_on:
|
|
service:
|
|
condition: service_started
|
|
```
|
|
|
|
### 2. No Hardcoded Networks
|
|
```yaml
|
|
# PROBLEM: overlap
|
|
networks:
|
|
gns-network:
|
|
ipam:
|
|
config:
|
|
- subnet: 172.28.0.0/16 # May conflict
|
|
|
|
# FIX: Docker auto-assigns
|
|
networks:
|
|
gns-network:
|
|
driver: bridge
|
|
```
|
|
|
|
### 3. Use Direct REST API
|
|
```typescript
|
|
// Direct REST client: fast, simple, no extra layers
|
|
const client = new GiteaClient({
|
|
apiUrl: config.giteaApiUrl,
|
|
token: config.giteaToken,
|
|
})
|
|
// All operations go directly to Gitea API via HTTP/REST
|
|
```
|
|
|
|
### 4. Pre-flight Validation
|
|
Before starting containers, validate prerequisites:
|
|
```bash
|
|
# Check if port is free, if not use another
|
|
curl -f http://localhost:3000/health || PORT=3001
|
|
|
|
# Check network doesn't exist
|
|
docker network ls | grep my-network && docker network rm my-network
|
|
|
|
# Check env vars are set
|
|
[ -z "$GITEA_TOKEN" ] && echo "WARNING: GITEA_TOKEN not set"
|
|
```
|
|
|
|
### 5. Self-Documenting Failures
|
|
If process must stop, write explicit "why" and "what to do" to both:
|
|
- Console output (human readable)
|
|
- Gitea issue comment (machine readable, includes `GNS_EVENT`)
|
|
|
|
```markdown
|
|
## 🚫 Agent Blocked
|
|
|
|
**Reason**: Gitea API not reachable
|
|
**Action**: Check `GITEA_API_URL` and `GITEA_TOKEN` environment variables
|
|
**Fallback**: Operations will use local file logging until API is available
|
|
```
|
|
|
|
## Implementation Checklist
|
|
|
|
For every new container/service:
|
|
- [ ] No `service_healthy` conditions in depends_on
|
|
- [ ] No hardcoded subnets or IPs
|
|
- [ ] Environment variables have safe fallbacks for startup
|
|
- [ ] Error boundaries in all async operations (try/catch)
|
|
- [ ] Error messages include both "what happened" and "next step"
|
|
- [ ] All operator-required steps are documented as checklist in issue body
|
|
|
|
## GNS-2 Event Format for Failures
|
|
|
|
```html
|
|
<!-- GNS_EVENT: {
|
|
"type": "system_failure",
|
|
"failure_point": "api_connection",
|
|
"requires_operator": true,
|
|
"reason": "GITEA_TOKEN not set, cannot connect to Gitea API",
|
|
"recovery_steps": [
|
|
"Set GITEA_TOKEN environment variable",
|
|
"Verify GITEA_API_URL is correct"
|
|
],
|
|
"fallback_active": "Local file logging",
|
|
"timestamp": "2026-05-08T22:23:00Z"
|
|
} -->
|
|
```
|
|
|
|
## Reference
|
|
- Docker compose depends_on behavior: https://docs.docker.com/compose/startup-order/
|
|
- Gitea API: `.kilo/shared/gitea-api.md`
|