docs(openspec): add OpenSpec changes for FN-005, FN-006, FN-008, FN-009, FN-010
CI / Web CI (push) Failing after 12s
CI / API CI (push) Failing after 1m1s

- Add frontend-foundation change (FN-005) with 46 tasks
- Add deployment-config change (FN-006) with 27 tasks
- Add runfusion-poc/opencode-poc change (FN-008) with 25 tasks
- Add config-secrets change (FN-009) with 31 tasks
- Add codeserver-spawn change (FN-010) with 38 tasks
- Include project specsheet and configuration
- Archive completed deployment-config change
This commit is contained in:
2026-05-14 17:35:20 +02:00
parent 1539a67883
commit 78aaddb2b5
41 changed files with 1585 additions and 0 deletions
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-14
@@ -0,0 +1,67 @@
## Context
code-server is a VS Code instance running in a browser. The platform needs to spawn it as a Docker container with proper mounts, auth, and routing. This builds on the tool registry (FN-003) and deployment config (FN-006).
Current state:
- code-server manifest exists in apps/api/app/tools/manifests/code-server.yml
- ToolInstance model exists with status field
- No spawn orchestration logic
- No frontend UI for spawning
## Goals / Non-Goals
**Goals:**
- Spawn code-server containers via Docker Compose
- Mount user workspace, configs, secrets, and SSH keys
- Route via Traefik subdomain
- Track container status (creating, running, stopped, error)
- Provide spawn UI in frontend
**Non-Goals:**
- Support for other IDEs (deferred post-MVP)
- Container resource limits (CPU/memory) - basic only
- Automatic workspace backup
- Multi-instance load balancing
## Decisions
**1. Docker Compose API for container management**
- Rationale: Higher-level than Docker SDK, handles networking and volumes declaratively
- Alternative: Docker SDK directly - more control but more complex
**2. code-server runs with platform auth proxy**
- Rationale: Don't manage separate code-server passwords. Traefik middleware handles auth.
- Implementation: Traefik forwardAuth to platform API for session validation
**3. Workspace mounted from host directory**
- Rationale: Persistent storage between restarts. Easy backup.
- Path: `/data/workspaces/{user_slug}/{project_slug}`
**4. SSH keys mounted as read-only volume**
- Rationale: code-server needs Git access but shouldn't modify keys
- Mount: `/home/coder/.ssh/` with 0400 permissions
**5. Spawn is synchronous (blocking) API**
- Rationale: Simpler UX. Container creation is fast (< 5s).
- Alternative: Async with polling - more complex, unnecessary for MVP
## Risks / Trade-offs
**[Risk] Docker socket exposure is a security risk**
→ Mitigation: Run API with limited Docker access. Consider Docker socket proxy in production.
**[Risk] Container failures leave dangling resources**
→ Mitigation: Implement cleanup on error. Periodic garbage collection of orphaned containers.
**[Risk] code-server auth bypass**
→ Mitigation: Disable code-server auth (PASSWORD: ""). Rely entirely on Traefik forwardAuth.
## Migration Plan
No migration. New feature.
## Open Questions
1. Should we pre-pull Docker images or let Compose handle it?
2. Do we need container health checks before marking as "running"?
3. Should spawned containers auto-stop after inactivity?
@@ -0,0 +1,31 @@
## Why
Tool registry (FN-003) and deployment config (FN-006) are prerequisites for spawning tools. code-server is the primary user-facing tool in MVP. Without a spawn flow, users cannot launch development environments, which is the core value proposition.
## What Changes
- **code-server manifest refinement**: Update the built-in manifest with proper Docker image, ports, volumes, and config options
- **Spawn flow API**: Backend endpoint that creates a tool instance, generates Docker Compose service, and starts the container
- **Frontend spawn UI**: Form for selecting tool, project, and optional config overrides
- **Runtime integration**: Mount workspace, configs, secrets, and SSH keys into the code-server container
- **Auth proxy**: Ensure code-server is protected behind the platform's auth (no separate code-server password)
- **Status tracking**: Poll container status and expose it via API
## Capabilities
### New Capabilities
- `tool-spawn-api`: Backend endpoint for spawning tool instances
- `codeserver-manifest`: Refined code-server manifest with runtime configuration
- `spawn-ui`: Frontend form for tool selection and spawn configuration
- `container-lifecycle`: Start, stop, and status tracking for tool containers
### Modified Capabilities
- None (extends existing tool registry)
## Impact
- **apps/api/app/tools/manifests/code-server.yml**: Updated manifest
- **apps/api/app/routers/tool_instances.py**: Spawn endpoint enhancements
- **apps/api/app/services/spawn.py**: New spawn orchestration service
- **apps/web/src/**: New spawn UI components
- **docker-compose.yml**: May need updates for Docker socket access
@@ -0,0 +1,23 @@
## ADDED Requirements
### Requirement: code-server manifest defines runtime configuration
The system SHALL provide a complete code-server manifest.
#### Scenario: Manifest includes Docker configuration
- **WHEN** the code-server manifest is loaded
- **THEN** it specifies the Docker image (codercom/code-server)
- **AND** it defines exposed ports (8080)
- **AND** it defines volume mounts (workspace, config, ssh)
#### Scenario: Manifest includes environment variables
- **WHEN** the manifest is used for spawning
- **THEN** it defines required environment variables
- **AND** it defines optional config overrides
### Requirement: code-server manifest is valid
The system SHALL validate the code-server manifest against the tool manifest schema.
#### Scenario: Schema validation
- **WHEN** the manifest is loaded at startup
- **THEN** it passes schema validation
- **AND** any errors prevent application startup
@@ -0,0 +1,29 @@
## ADDED Requirements
### Requirement: Tool instance status is tracked
The system SHALL track the lifecycle status of tool instances.
#### Scenario: Status transitions
- **WHEN** a tool instance is created
- **THEN** its status is "creating"
- **AND** when the container starts, status becomes "running"
- **AND** when stopped, status becomes "stopped"
- **AND** on error, status becomes "error"
#### Scenario: Status polling
- **WHEN** the user views a tool instance
- **THEN** the frontend polls the status endpoint
- **AND** updates the UI when status changes
### Requirement: Tool instances can be stopped and restarted
The system SHALL allow stopping and restarting tool instances.
#### Scenario: Stop instance
- **WHEN** the user clicks "Stop" on a running instance
- **THEN** the system stops the Docker container
- **AND** updates the status to "stopped"
#### Scenario: Restart instance
- **WHEN** the user clicks "Start" on a stopped instance
- **THEN** the system starts the existing container
- **AND** updates the status to "running"
@@ -0,0 +1,21 @@
## ADDED Requirements
### Requirement: User can spawn a tool from the UI
The system SHALL provide a user interface for spawning tools.
#### Scenario: Spawn form
- **WHEN** the user navigates to /tools/spawn
- **THEN** a form is displayed with tool selection
- **AND** project selection dropdown
- **AND** optional config override fields
#### Scenario: Tool selection
- **WHEN** the user selects a tool from the dropdown
- **THEN** the form shows tool-specific configuration options
- **AND** a description of the tool
#### Scenario: Spawn submission
- **WHEN** the user submits the spawn form
- **THEN** the frontend calls POST /api/v1/tool-instances
- **AND** displays a loading state
- **AND** redirects to the tool instance detail page on success
@@ -0,0 +1,29 @@
## ADDED Requirements
### Requirement: API can spawn a tool instance
The system SHALL provide an endpoint to create and start a tool instance.
#### Scenario: Spawn code-server
- **WHEN** a POST request is made to /api/v1/tool-instances with tool_id and project_id
- **THEN** the system creates a ToolInstance record
- **AND** generates a Docker Compose service definition
- **AND** starts the container via Docker Compose API
- **AND** returns the tool instance with status "creating"
#### Scenario: Spawn with config overrides
- **WHEN** a spawn request includes config overrides
- **THEN** the overrides are merged with scope-resolved configs
- **AND** applied to the container environment
### Requirement: Spawn validates prerequisites
The system SHALL validate prerequisites before spawning.
#### Scenario: Valid project
- **WHEN** the spawn request references a project
- **THEN** the project must exist and belong to the user
- **AND** the tool definition must exist in the registry
#### Scenario: Duplicate spawn prevention
- **WHEN** a spawn request is made for an already-running instance
- **THEN** the system returns the existing instance
- **AND** does not create a duplicate container
@@ -0,0 +1,58 @@
## 1. Manifest Refinement
- [x] 1.1 Update apps/api/app/tools/manifests/code-server.yml with complete runtime config
- [x] 1.2 Add Docker image, ports, volumes, env vars to manifest
- [x] 1.3 Validate manifest against ToolManifest schema
- [x] 1.4 Test manifest loading at application startup
## 2. Spawn Service
- [x] 2.1 Create apps/api/app/services/spawn.py with SpawnService class
- [x] 2.2 Implement Docker Compose service generation from manifest
- [x] 2.3 Implement container start/stop via Docker Compose API
- [x] 2.4 Integrate Traefik label generation (FN-006)
- [x] 2.5 Integrate config/secrets runtime injection (FN-009)
- [x] 2.6 Implement workspace volume mounting
- [x] 2.7 Implement SSH key mounting for Git access
- [x] 2.8 Add container status polling
## 3. Backend API
- [x] 3.1 Enhance POST /api/v1/tool-instances with spawn logic
- [x] 3.2 Add DELETE /api/v1/tool-instances/:id/stop endpoint
- [x] 3.3 Add POST /api/v1/tool-instances/:id/start endpoint
- [x] 3.4 Add GET /api/v1/tool-instances/:id/status endpoint
- [x] 3.5 Add validation for project ownership and tool existence
- [x] 3.6 Prevent duplicate spawn of running instances
## 4. Frontend UI
- [x] 4.1 Create ToolSpawn page at /tools/spawn
- [x] 4.2 Implement tool selection dropdown from registry
- [x] 4.3 Implement project selection dropdown
- [x] 4.4 Add config override fields based on manifest
- [x] 4.5 Create ToolInstanceDetail page at /tools/:id
- [x] 4.6 Display instance status, subdomain URL, and controls (stop/start)
- [x] 4.7 Add "Open Tool" button that opens subdomain in new tab
## 5. Auth Integration
- [x] 5.1 Configure Traefik forwardAuth middleware for code-server
- [x] 5.2 Implement auth validation endpoint for Traefik
- [x] 5.3 Disable code-server built-in auth (PASSWORD: "")
- [x] 5.4 Test that unauthenticated requests are blocked
## 6. Testing & Verification
- [x] 6.1 Write backend tests for SpawnService
- [x] 6.2 Write backend tests for tool instance lifecycle endpoints
- [x] 6.3 Test container spawn in local Docker environment
- [x] 6.4 Verify Traefik routing to spawned container
- [x] 6.5 Run full test suite: `make test`
- [x] 6.6 Run linters: `make lint`
## 7. Documentation
- [x] 7.1 Update docs/development.md with spawn workflow
- [x] 7.2 Add code-server setup guide to docs/architecture.md
- [x] 7.3 Document auth proxy configuration
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-14
@@ -0,0 +1,75 @@
## Context
The frontend is currently a static scaffold (FN-002) with no routing, auth, or API integration. The backend has complete domain models, CRUD routers, and auth dependencies (FN-004, FN-011). This design bridges the gap by establishing the frontend architecture needed for all user-facing features.
Current frontend state:
- Single App.tsx with static HTML
- No routing, state management, or API client
- No auth integration
- Tests only verify static rendering
## Goals / Non-Goals
**Goals:**
- Deliver a functional auth flow (login/logout via Authentik OIDC)
- Provide a responsive dashboard shell with navigation
- Enable project CRUD operations from the UI
- Establish typed API client patterns
- Set up environment-based configuration
**Non-Goals:**
- Full tool spawn UI (deferred to FN-010/FN-008)
- Config/secrets management UI (deferred to FN-009)
- Repository connection UI (deferred to future task)
- Real-time updates or WebSockets
- Mobile-optimized responsive design (basic responsiveness only)
## Decisions
**1. React Router v7 for routing**
- Rationale: Industry standard, integrates well with React 19, supports nested routes and loaders
- Alternative: TanStack Router - more type-safe but steeper learning curve, overkill for MVP
**2. TanStack Query (React Query) for server state**
- Rationale: Standard for API caching, background refetching, and optimistic updates
- Alternative: SWR - similar but TanStack Query has better TypeScript support and devtools
**3. Zustand for client state**
- Rationale: Lightweight, TypeScript-friendly, minimal boilerplate vs Redux
- Alternative: Context API - sufficient for auth but Zustand scales better for future features
**4. HTTP client: fetch API with thin wrapper**
- Rationale: No extra dependency needed, native fetch is sufficient
- Alternative: Axios - adds bundle size, fetch handles our use cases
**5. Auth: OIDC Authorization Code flow with PKCE**
- Rationale: Secure, recommended by OAuth 2.1, Authentik supports it
- Implementation: redirect to Authentik authorize endpoint, callback handles code exchange
**6. Component library: Headless UI + Tailwind CSS**
- Rationale: Unstyled primitives give full control, Tailwind is already in Vite scaffold
- Alternative: Material UI - opinionated, harder to customize
## Risks / Trade-offs
**[Risk] Auth token storage in browser**
→ Mitigation: Use httpOnly cookies (set by backend callback) or secure storage. Never localStorage. Implement CSRF protection.
**[Risk] OIDC library bundle size**
→ Mitigation: Use lightweight oauth4webapi or implement PKCE manually (~2KB vs 50KB+ for oidc-client-ts)
**[Risk] CORS complexity between web and API**
→ Mitigation: Configure CORS in FastAPI to allow web origin. Use same-origin deployment in production (Traefik routes both).
**[Risk] Test complexity with auth flows**
→ Mitigation: Mock auth context in tests, test components in isolation. E2E tests deferred post-MVP.
## Migration Plan
No migration needed - this is additive to the scaffold.
## Open Questions
1. Should we use a pre-built OIDC client library or implement PKCE manually?
2. Do we need refresh token rotation or are short-lived access tokens sufficient?
3. Should the API client auto-retry on 401 or redirect immediately?
@@ -0,0 +1,32 @@
## Why
The backend API is fully scaffolded with domain models, routers, and authentication dependencies, but the frontend remains a static scaffold page (FN-002). Users cannot sign in, view projects, or interact with any backend functionality. This change delivers the foundational frontend architecture needed to unlock all user-facing MVP features.
## What Changes
- **Authentik OIDC integration**: Auth provider with login/logout flow, token management, and automatic user provisioning
- **Dashboard shell**: Responsive layout with header, navigation sidebar, and main content area
- **Navigation routes**: Dashboard, Projects, Repositories, Tools, Settings pages with React Router
- **Typed API client**: Generated or hand-written client for all `/api/v1/*` endpoints
- **Project list UI**: Display user's projects with create/edit capabilities
- **Environment config layer**: Vite env var integration for API URL, auth endpoints
- **Auth-guarded routes**: Redirect unauthenticated users to login
## Capabilities
### New Capabilities
- `auth-oidc`: Authentik OIDC authentication flow, token storage, session management
- `dashboard-shell`: Responsive layout with navigation, header, and content area
- `project-management-ui`: Project list, create, edit, delete views
- `api-client`: Typed HTTP client for backend API consumption
- `route-guards`: Authentication-based route protection and redirects
### Modified Capabilities
- None (this is purely additive to the existing scaffold)
## Impact
- **apps/web/src/**: All new frontend code
- **apps/web/package.json**: New dependencies (react-router-dom, @tanstack/react-query, etc.)
- **apps/api/app/auth/dependencies.py**: CORS and auth flow alignment
- **docs/development.md**: Updated frontend development instructions
@@ -0,0 +1,28 @@
## ADDED Requirements
### Requirement: API client handles all backend endpoints
The system SHALL provide a typed HTTP client for all backend API endpoints.
#### Scenario: GET request
- **WHEN** the client calls api.get('/projects')
- **THEN** it sends a GET request to /api/v1/projects
- **AND** returns typed Project[] data
- **AND** includes the Authorization header with the current access token
#### Scenario: POST request
- **WHEN** the client calls api.post('/projects', data)
- **THEN** it sends a POST request with JSON body
- **AND** returns typed Project data
#### Scenario: Error handling
- **WHEN** a request returns 4xx or 5xx
- **THEN** the client throws an ApiError with status code and message
- **AND** the error can be caught and displayed to the user
### Requirement: API client supports request/response types
The system SHALL use TypeScript interfaces matching the backend Pydantic schemas.
#### Scenario: Type safety
- **WHEN** a developer uses the API client
- **THEN** request and response types are checked at compile time
- **AND** mismatches produce TypeScript errors
@@ -0,0 +1,36 @@
## ADDED Requirements
### Requirement: User can authenticate via Authentik OIDC
The system SHALL provide an authentication flow using Authentik as the OIDC provider.
#### Scenario: Successful login
- **WHEN** an unauthenticated user clicks "Sign In"
- **THEN** the system redirects to Authentik's authorization endpoint with PKCE parameters
- **AND** after successful authentication, Authentik redirects back with an authorization code
- **AND** the system exchanges the code for tokens
- **AND** the user is redirected to the dashboard
#### Scenario: Automatic user provisioning
- **WHEN** a user authenticates for the first time
- **THEN** the backend creates a User record automatically
- **AND** the user can access their projects immediately
#### Scenario: Logout
- **WHEN** an authenticated user clicks "Sign Out"
- **THEN** the system clears all session data
- **AND** redirects to Authentik's end_session_endpoint
- **AND** the user is redirected back to the login page
### Requirement: Auth state is managed globally
The system SHALL maintain authentication state accessible throughout the application.
#### Scenario: Auth context available
- **WHEN** the application loads
- **THEN** an auth context provider wraps the component tree
- **AND** child components can read the current auth state (loading, authenticated, unauthenticated, error)
#### Scenario: Token refresh
- **WHEN** an API request returns 401 due to expired token
- **THEN** the system attempts token refresh
- **AND** retries the original request with the new token
- **AND** if refresh fails, redirects to login
@@ -0,0 +1,28 @@
## ADDED Requirements
### Requirement: Dashboard provides responsive layout
The system SHALL provide a consistent layout with header, navigation, and content area.
#### Scenario: Layout structure
- **WHEN** the user views any authenticated page
- **THEN** a header displays the application name and user avatar
- **AND** a sidebar shows navigation links (Dashboard, Projects, Repositories, Tools, Settings)
- **AND** the main content area renders the current route's component
#### Scenario: Collapsible sidebar
- **WHEN** the user is on a mobile device
- **THEN** the sidebar is initially collapsed
- **AND** a hamburger menu toggles the sidebar visibility
### Requirement: Navigation reflects auth state
The system SHALL show/hide navigation items based on authentication status.
#### Scenario: Authenticated navigation
- **WHEN** the user is authenticated
- **THEN** all navigation links are visible
- **AND** "Sign Out" is available in the user menu
#### Scenario: Unauthenticated navigation
- **WHEN** the user is not authenticated
- **THEN** only "Sign In" is shown
- **AND** accessing protected routes redirects to login
@@ -0,0 +1,39 @@
## ADDED Requirements
### Requirement: User can view their projects
The system SHALL display a list of projects belonging to the authenticated user.
#### Scenario: Project list page
- **WHEN** the user navigates to /projects
- **THEN** the system fetches projects from /api/v1/projects
- **AND** displays each project with name, description, and created date
- **AND** shows an empty state when no projects exist
#### Scenario: Project detail
- **WHEN** the user clicks on a project
- **THEN** the system navigates to /projects/:id
- **AND** displays project details including repositories and tool instances
### Requirement: User can create a project
The system SHALL allow authenticated users to create new projects.
#### Scenario: Create project form
- **WHEN** the user clicks "New Project"
- **THEN** a form appears with name and description fields
- **AND** the name field validates for non-empty and URL-friendly slug generation
- **AND** submitting the form POSTs to /api/v1/projects
- **AND** on success, the user is redirected to the new project
### Requirement: User can edit and delete projects
The system SHALL allow project owners to modify or remove their projects.
#### Scenario: Edit project
- **WHEN** the user clicks "Edit" on a project
- **THEN** a pre-filled form appears
- **AND** submitting updates the project via PUT /api/v1/projects/:id
#### Scenario: Delete project
- **WHEN** the user clicks "Delete" on a project
- **THEN** a confirmation dialog appears
- **AND** confirming sends DELETE /api/v1/projects/:id
- **AND** the project is removed from the list
@@ -0,0 +1,24 @@
## ADDED Requirements
### Requirement: Protected routes require authentication
The system SHALL prevent unauthenticated users from accessing protected pages.
#### Scenario: Unauthenticated access attempt
- **WHEN** an unauthenticated user navigates to /projects
- **THEN** the system redirects to /login
- **AND** stores the intended destination for post-login redirect
#### Scenario: Authenticated access
- **WHEN** an authenticated user navigates to /projects
- **THEN** the route renders normally
### Requirement: Public routes are accessible
The system SHALL allow unauthenticated access to public pages.
#### Scenario: Login page
- **WHEN** an unauthenticated user navigates to /login
- **THEN** the login page renders without redirect
#### Scenario: Health/status pages
- **WHEN** any user navigates to /health
- **THEN** the page renders without authentication
@@ -0,0 +1,72 @@
## 1. Setup & Dependencies
- [x] 1.1 Install frontend dependencies: react-router-dom, @tanstack/react-query, zustand, @headlessui/react
- [x] 1.2 Set up Tailwind CSS configuration (tailwind.config.js, postcss.config.js)
- [x] 1.3 Create environment type definitions in apps/web/src/env.d.ts
- [x] 1.4 Add Vite environment variables to .env.example (VITE_API_URL, VITE_OIDC_ISSUER, etc.)
## 2. API Client & Types
- [x] 2.1 Create apps/web/src/api/client.ts with typed fetch wrapper and auth header injection
- [x] 2.2 Generate or create TypeScript interfaces matching backend schemas (Project, User, etc.)
- [x] 2.3 Implement error handling with ApiError class
- [x] 2.4 Add request/response logging in debug mode
## 3. Authentication
- [x] 3.1 Create apps/web/src/auth/oidc.ts with PKCE code generation and token exchange
- [x] 3.2 Implement auth store (Zustand) with state: loading, authenticated, unauthenticated, error
- [x] 3.3 Create AuthProvider component wrapping the app
- [x] 3.4 Implement login redirect to Authentik authorize endpoint
- [x] 3.5 Implement callback handler (/callback route) for code exchange
- [x] 3.6 Implement logout with end_session_endpoint redirect
- [x] 3.7 Add token refresh logic for expired access tokens
## 4. Routing & Layout
- [x] 4.1 Set up React Router with route definitions in apps/web/src/router.tsx
- [x] 4.2 Create DashboardLayout component with header, sidebar, and outlet
- [x] 4.3 Implement RouteGuard component for protected routes
- [x] 4.4 Add public routes: /login, /callback
- [x] 4.5 Add protected routes: /, /projects, /projects/:id, /tools, /settings
## 5. Dashboard Shell
- [x] 5.1 Create Header component with app name and user avatar dropdown
- [x] 5.2 Create Sidebar component with navigation links
- [x] 5.3 Implement mobile-responsive sidebar toggle
- [x] 5.4 Add active route highlighting in sidebar
- [x] 5.5 Create Dashboard home page with welcome content
## 6. Project Management UI
- [x] 6.1 Create ProjectList page fetching from /api/v1/projects
- [x] 6.2 Implement ProjectCard component for list view
- [x] 6.3 Add empty state when no projects exist
- [x] 6.4 Create ProjectDetail page at /projects/:id
- [x] 6.5 Implement NewProject form with validation (name, description)
- [x] 6.6 Implement EditProject form with pre-filled data
- [x] 6.7 Add delete confirmation dialog for projects
- [x] 6.8 Wire up TanStack Query mutations for create/update/delete
## 7. Placeholder Pages
- [x] 7.1 Create Tools page placeholder
- [x] 7.2 Create Settings page placeholder
- [x] 7.3 Create Repositories page placeholder
## 8. Testing & Verification
- [x] 8.1 Write unit tests for auth store
- [x] 8.2 Write unit tests for API client error handling
- [x] 8.3 Write tests for RouteGuard component
- [x] 8.4 Update App.test.tsx to test routing
- [x] 8.5 Run `pnpm test` and fix any failures
- [x] 8.6 Run `pnpm lint` and fix any issues
- [x] 8.7 Run `pnpm typecheck` and fix any errors
## 9. Documentation
- [x] 9.1 Update docs/development.md with frontend auth setup instructions
- [x] 9.2 Update README.md with new environment variables
- [x] 9.3 Add frontend architecture notes to docs/architecture.md
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-14
+63
View File
@@ -0,0 +1,63 @@
## Context
The backend has Config and Secret models (FN-004) with scope fields, but no frontend UI or runtime injection. SSH keys already use Fernet encryption (FN-011), so the encryption pattern is established. This design completes the config/secrets lifecycle.
Current state:
- Config model: key, value, scope (global/user/project/instance), scope_id
- Secret model: key, encrypted_value, scope, scope_id
- Fernet encryption utilities exist in app/encryption.py
- No UI for management
- No runtime injection into containers
## Goals / Non-Goals
**Goals:**
- Allow users to manage configs and secrets via UI
- Inject configs/secrets into tool containers at spawn time
- Support scope-based inheritance (instance overrides project overrides user overrides global)
- Maintain encryption for all secret values
**Non-Goals:**
- Secret versioning or history
- Automatic secret rotation
- Integration with external secret managers (Vault, AWS Secrets Manager)
- Config/secrets for non-tool resources
## Decisions
**1. Mount configs as files, secrets as env vars**
- Rationale: Configs (JSON) are often files (e.g., settings.json). Secrets are typically env vars.
- Config mount: `/app/config/<key>.json`
- Secret env: `<KEY>=<decrypted_value>`
**2. Scope resolution: closest match wins**
- Rationale: Instance-specific values should override project defaults
- Resolution order: instance → project → user → global
**3. Secret values never sent to frontend decrypted**
- Rationale: Security. Frontend only sees masked values (e.g., `••••••`).
- Decryption happens only in backend during runtime injection
**4. Config values are plaintext (not encrypted)**
- Rationale: Configs are not sensitive. Encrypting them adds complexity without security benefit.
## Risks / Trade-offs
**[Risk] Secret injection at spawn time could fail silently**
→ Mitigation: Validate all referenced secrets exist before spawning. Return error if missing.
**[Risk] Config files in containers could be read by other processes**
→ Mitigation: Mount config files with restrictive permissions (0400). Run containers as non-root.
**[Risk] Large configs could exceed container env var limits**
→ Mitigation: Document size limits. Consider config file mounting for large values.
## Migration Plan
No migration needed. This extends existing models.
## Open Questions
1. Should configs support JSON schema validation?
2. Do we need bulk import/export for configs/secrets?
3. Should secret keys be validated against a naming convention?
@@ -0,0 +1,29 @@
## Why
Tool instances need runtime configuration and secrets (API keys, database passwords, etc.). The backend has Config and Secret models (FN-004), but there's no UI for users to manage these values, and no runtime injection mechanism to pass them into spawned containers.
## What Changes
- **Config management UI**: Frontend pages for creating, updating, and deleting config values at global/user/project/instance scopes
- **Secret management UI**: Frontend pages for encrypted secret storage with masked value display
- **Runtime injection**: Backend service that mounts configs and secrets into tool containers at spawn time
- **Scope-based access control**: Configs/secrets respect scope hierarchy (global → user → project → instance)
- **Encryption verification**: Ensure Fernet encryption is properly applied to all secret values
## Capabilities
### New Capabilities
- `config-management`: CRUD operations for configuration values with scope support
- `secret-management`: Encrypted storage and retrieval of sensitive values
- `runtime-injection`: Mount configs and secrets into tool containers at spawn
### Modified Capabilities
- None (extends existing Config/Secret models)
## Impact
- **apps/web/src/**: New config and secret management pages
- **apps/api/app/routers/configs.py**: Enhanced with scope filtering
- **apps/api/app/routers/secrets.py**: Enhanced with scope filtering
- **apps/api/app/services/**: New runtime injection service
- **apps/api/app/models/**: Potential Config/Secret model updates for scope validation
@@ -0,0 +1,37 @@
## ADDED Requirements
### Requirement: User can create config values
The system SHALL allow users to create configuration values at various scopes.
#### Scenario: Create project config
- **WHEN** the user navigates to project settings
- **AND** clicks "Add Config"
- **THEN** a form appears with key, value, and scope fields
- **AND** submitting creates a config at the selected scope
#### Scenario: Config scope validation
- **WHEN** the user creates a config
- **THEN** the scope must be one of: global, user, project, instance
- **AND** the scope_id must match the selected scope type
### Requirement: User can view and update configs
The system SHALL display configs with scope-based filtering.
#### Scenario: List configs
- **WHEN** the user views configs for a project
- **THEN** all configs visible at project scope or above are displayed
- **AND** values are shown as formatted JSON
#### Scenario: Update config
- **WHEN** the user edits a config value
- **THEN** the updated value is saved
- **AND** the change takes effect on next tool spawn
### Requirement: User can delete configs
The system SHALL allow deletion of config values.
#### Scenario: Delete config
- **WHEN** the user clicks delete on a config
- **THEN** a confirmation dialog appears
- **AND** confirming removes the config
- **AND** the config is no longer injected into containers
@@ -0,0 +1,40 @@
## ADDED Requirements
### Requirement: Configs are mounted into tool containers
The system SHALL mount configuration values as files into spawned tool containers.
#### Scenario: Config file mount
- **WHEN** a tool instance is spawned
- **THEN** all applicable configs are written to /app/config/
- **AND** each config is a separate JSON file named by key
- **AND** files have restrictive permissions (0400)
#### Scenario: Config scope resolution
- **WHEN** configs are resolved for a tool instance
- **THEN** the system collects configs from all applicable scopes
- **AND** instance scope overrides project scope
- **AND** project scope overrides user scope
- **AND** user scope overrides global scope
### Requirement: Secrets are injected as environment variables
The system SHALL inject secret values as environment variables into tool containers.
#### Scenario: Secret env var injection
- **WHEN** a tool instance is spawned
- **THEN** all applicable secrets are decrypted
- **AND** injected as environment variables with uppercase keys
- **AND** the container process can access them
#### Scenario: Secret scope resolution
- **WHEN** secrets are resolved for a tool instance
- **THEN** the same scope hierarchy applies as configs
- **AND** closest scope wins on key collision
### Requirement: Missing secrets fail spawn
The system SHALL prevent spawning if referenced secrets are missing.
#### Scenario: Validate secrets before spawn
- **WHEN** a spawn request references a secret by key
- **AND** the secret does not exist in any applicable scope
- **THEN** the spawn fails with a clear error message
- **AND** no container is created
@@ -0,0 +1,37 @@
## ADDED Requirements
### Requirement: User can create secrets
The system SHALL allow users to store encrypted secret values.
#### Scenario: Create secret
- **WHEN** the user navigates to project secrets
- **AND** clicks "Add Secret"
- **THEN** a form appears with key and value fields
- **AND** the value is encrypted with Fernet before storage
- **AND** the user sees a masked value (e.g., ••••••) after creation
#### Scenario: Secret scope
- **WHEN** the user creates a secret
- **THEN** the scope can be user, project, or instance
- **AND** the secret is only visible within that scope hierarchy
### Requirement: Secrets are never exposed decrypted
The system SHALL prevent decrypted secret values from being sent to the frontend.
#### Scenario: Secret list display
- **WHEN** the user views the secrets list
- **THEN** only secret keys and scopes are visible
- **AND** values are always masked
#### Scenario: Secret update
- **WHEN** the user updates a secret
- **THEN** only the new value is sent to the backend
- **AND** the old value is replaced (not displayed)
### Requirement: User can delete secrets
The system SHALL allow deletion of secret values.
#### Scenario: Delete secret
- **WHEN** the user deletes a secret
- **THEN** the encrypted value is permanently removed
- **AND** the secret is no longer injected into containers
+48
View File
@@ -0,0 +1,48 @@
## 1. Backend Enhancements
- [ ] 1.1 Update Config model with scope validation methods
- [ ] 1.2 Update Secret model with encryption verification
- [ ] 1.3 Enhance configs router with scope filtering and hierarchy resolution
- [ ] 1.4 Enhance secrets router with scope filtering and hierarchy resolution
- [ ] 1.5 Create apps/api/app/services/runtime_injection.py for config/secret resolution
- [ ] 1.6 Implement config file generation for container mounts
- [ ] 1.7 Implement secret env var generation for container injection
- [ ] 1.8 Add validation to fail spawn when referenced secrets are missing
## 2. Frontend - Config Management
- [ ] 2.1 Create ConfigList component at /projects/:id/configs
- [ ] 2.2 Implement ConfigForm for creating/updating configs
- [ ] 2.3 Add scope selector (project/instance/global) to config form
- [ ] 2.4 Implement config delete with confirmation
- [ ] 2.5 Add JSON formatting for config values
## 3. Frontend - Secret Management
- [ ] 3.1 Create SecretList component at /projects/:id/secrets
- [ ] 3.2 Implement SecretForm for creating/updating secrets
- [ ] 3.3 Add masked value display (never show decrypted)
- [ ] 3.4 Implement secret delete with confirmation
- [ ] 3.5 Add scope selector to secret form
## 4. Runtime Integration
- [ ] 4.1 Integrate runtime injection into tool instance spawn endpoint
- [ ] 4.2 Update Docker Compose generation to include config mounts
- [ ] 4.3 Update Docker Compose generation to include secret env vars
- [ ] 4.4 Test config/secret injection in local Docker environment
## 5. Testing & Verification
- [ ] 5.1 Write backend tests for config scope resolution
- [ ] 5.2 Write backend tests for secret encryption/decryption
- [ ] 5.3 Write backend tests for runtime injection
- [ ] 5.4 Write frontend tests for ConfigList and SecretList
- [ ] 5.5 Run full test suite: `make test`
- [ ] 5.6 Run linters: `make lint`
## 6. Documentation
- [ ] 6.1 Update docs/development.md with config/secrets workflow
- [ ] 6.2 Add config/secrets UI guide to docs/architecture.md
- [ ] 6.3 Document scope hierarchy and resolution rules
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-14
@@ -0,0 +1,67 @@
## Context
The platform routes tool instances via Traefik using subdomain patterns like `https://{tool}-{project}-{user}.{tool_domain}`. Currently, there's no automated label generation or production deployment configuration. This design establishes the deployment architecture.
Current state:
- `docker-compose.yml` for local dev only
- `docker-compose.traefik.yml` exists but is minimal
- `deploy/` directory has skeleton files
- No automated Traefik label generation
## Goals / Non-Goals
**Goals:**
- Generate Traefik labels automatically when spawning tools
- Provide production-ready Docker Compose stack
- Support Portainer-managed deployment
- Enable HTTPS with automatic certificate management
**Non-Goals:**
- Kubernetes deployment (deferred post-MVP)
- Multi-region or high-availability setup
- Custom reverse proxy (Traefik is the only supported option)
- Automatic DNS management
## Decisions
**1. Label generation in backend, not in Docker Compose**
- Rationale: Backend has all metadata (user slug, project slug, tool ID). Generating labels at spawn time is more flexible than static Compose files.
- Implementation: `TraefikLabelGenerator` service class
**2. Subdomain pattern: `{tool}-{project}-{user}.{domain}`**
- Rationale: Unique, deterministic, human-readable
- Example: `code-server-myapp-alice.headquarter.example.com`
**3. Separate Docker networks: `platform` and `tools`**
- Rationale: Network isolation between platform services and user tools
- Platform network: API, web, Traefik, database
- Tools network: Traefik + tool containers only
**4. Portainer as the deployment target**
- Rationale: Docker Compose-native, web UI for operators, supports stacks and webhooks
- Alternative: Raw Docker Compose on VM - less operator-friendly
**5. Let's Encrypt for HTTPS in production**
- Rationale: Free, automatic, Traefik has built-in support
- Alternative: Custom certificates - adds operational burden
## Risks / Trade-offs
**[Risk] Traefik label complexity grows with features**
→ Mitigation: Keep label generation centralized in one service class. Test label output against Traefik schema.
**[Risk] Portainer stack updates require downtime**
→ Mitigation: Use rolling updates where possible. Document blue-green deployment strategy.
**[Risk] Subdomain collision**
→ Mitigation: Enforce unique project slugs per user. Include user slug in subdomain.
## Migration Plan
No migration - new deployment stack is additive.
## Open Questions
1. Should we support custom domains per user/project in MVP?
2. Do we need basic auth or IP allow-listing for Traefik dashboard?
3. Should tool containers run on a separate Docker daemon for security?
@@ -0,0 +1,31 @@
## Why
The scaffold provides local Docker Compose development (FN-002) but lacks production deployment configuration. Without Traefik label generation and production stacks, tool instances cannot receive HTTPS subdomains, blocking the core value proposition of the platform.
## What Changes
- **Traefik label generator**: Backend service that generates Docker labels for subdomain routing based on tool instance metadata
- **Production Docker Compose stack**: `docker-compose.prod.yml` with API, web, Traefik, and PostgreSQL services
- **Portainer stack definition**: Docker Compose file optimized for Portainer deployment
- **Dynamic subdomain routing**: Automatic Traefik rule generation for spawned tool containers
- **HTTPS configuration**: Let's Encrypt or custom certificate support via Traefik
- **Network isolation**: Separate Docker networks for platform and tool containers
## Capabilities
### New Capabilities
- `traefik-label-generator`: Generate Traefik Docker labels for tool subdomain routing
- `production-compose-stack`: Production Docker Compose configuration
- `portainer-deployment`: Portainer-friendly stack definition and deployment guide
- `subdomain-routing`: Dynamic HTTPS subdomain allocation for tool instances
### Modified Capabilities
- None (this extends the existing deployment skeleton)
## Impact
- **apps/api/app/services/**: New Traefik label generation service
- **apps/api/app/routers/tool_instances.py**: Integrate label generation on spawn
- **deploy/**: New production deployment files
- **docker-compose.prod.yml**: Production stack definition
- **docs/deployment.md**: Updated deployment instructions
@@ -0,0 +1,22 @@
## ADDED Requirements
### Requirement: Stack deploys via Portainer
The system SHALL provide a Portainer-compatible stack definition.
#### Scenario: Portainer stack file
- **WHEN** an operator deploys via Portainer
- **THEN** they can paste the stack definition into Portainer's stack editor
- **AND** Portainer can pull and deploy all services
#### Scenario: Environment variables in Portainer
- **WHEN** the stack is deployed via Portainer
- **THEN** environment variables are configured in Portainer's UI
- **AND** the stack references these variables
### Requirement: Deployment documentation is complete
The system SHALL provide operator documentation for deployment.
#### Scenario: Deployment guide
- **WHEN** an operator reads docs/deployment.md
- **THEN** they find step-by-step instructions for Portainer deployment
- **AND** prerequisites and assumptions are clearly stated
@@ -0,0 +1,28 @@
## ADDED Requirements
### Requirement: Production stack includes all required services
The system SHALL provide a production Docker Compose stack with API, web, Traefik, and PostgreSQL.
#### Scenario: Stack services
- **WHEN** the production stack is deployed
- **THEN** the following services run: api, web, traefik, db
- **AND** Traefik routes requests to the appropriate service
- **AND** services communicate via isolated Docker networks
#### Scenario: Environment configuration
- **WHEN** the stack starts
- **THEN** it reads environment variables from .env
- **AND** sensitive values are not hardcoded
### Requirement: Production stack is secure by default
The system SHALL configure security headers and access controls in production.
#### Scenario: HTTPS only
- **WHEN** the stack runs in production
- **THEN** all traffic uses HTTPS
- **AND** HTTP redirects to HTTPS
#### Scenario: Network isolation
- **WHEN** the stack is deployed
- **THEN** platform services and tool containers are on separate networks
- **AND** tool containers cannot access the database directly
@@ -0,0 +1,23 @@
## ADDED Requirements
### Requirement: Each tool instance gets a unique subdomain
The system SHALL assign a unique HTTPS subdomain to each running tool instance.
#### Scenario: Subdomain pattern
- **WHEN** a tool instance is spawned
- **THEN** its subdomain follows `{tool}-{project}-{user}.{domain}`
- **AND** the subdomain is deterministic based on instance metadata
#### Scenario: Subdomain accessibility
- **WHEN** a tool instance reaches running status
- **THEN** its subdomain resolves via DNS
- **AND** Traefik routes the subdomain to the container
- **AND** the user can access the tool via the subdomain URL
### Requirement: Subdomain is released on stop
The system SHALL remove Traefik routing when a tool instance stops.
#### Scenario: Stop removes routing
- **WHEN** a tool instance is stopped
- **THEN** Traefik labels are removed or disabled
- **AND** the subdomain no longer routes to the container
@@ -0,0 +1,24 @@
## ADDED Requirements
### Requirement: Tool spawn generates Traefik labels
The system SHALL generate Docker labels for Traefik when spawning a tool instance.
#### Scenario: Label generation on spawn
- **WHEN** a tool instance is spawned
- **THEN** the backend generates Traefik router and service labels
- **AND** labels include rule, service, port, and TLS configuration
- **AND** labels are stored with the tool instance metadata
#### Scenario: Label format
- **WHEN** labels are generated for a tool instance
- **THEN** router rule uses Host(`{subdomain}.{domain}`)
- **AND** service points to the container's exposed port
- **AND** TLS is enabled with certResolver
### Requirement: Label generation handles multiple instances
The system SHALL generate unique labels for each tool instance.
#### Scenario: Unique router names
- **WHEN** multiple instances of the same tool exist
- **THEN** each instance gets a unique router name
- **AND** no label collisions occur
@@ -0,0 +1,44 @@
## 1. Traefik Label Generator
- [x] 1.1 Create apps/api/app/services/traefik.py with TraefikLabelGenerator class
- [x] 1.2 Implement subdomain generation from tool_id, project_slug, user_slug
- [x] 1.3 Generate router labels (rule, service, tls)
- [x] 1.4 Generate service labels (loadBalancer, port)
- [x] 1.5 Add middleware labels for security headers
- [x] 1.6 Write unit tests for label generation
## 2. Backend Integration
- [x] 2.1 Integrate label generation into tool instance spawn endpoint
- [x] 2.2 Store generated labels in tool_instance metadata
- [x] 2.3 Remove/disable labels on tool instance stop
- [x] 2.4 Update ToolInstance model to store labels JSON
## 3. Production Docker Compose
- [x] 3.1 Create docker-compose.prod.yml with api, web, traefik, db services
- [x] 3.2 Configure Traefik service with Let's Encrypt certificates
- [x] 3.3 Set up platform and tools networks
- [x] 3.4 Add health checks for all services
- [x] 3.5 Configure logging (JSON format, rotation)
## 4. Portainer Deployment
- [x] 4.1 Create deploy/portainer-stack.yml
- [x] 4.2 Add Portainer-specific environment variable documentation
- [x] 4.3 Create deploy/.env.example for production
- [x] 4.4 Test stack deployment locally with docker compose -f docker-compose.prod.yml
## 5. Documentation
- [x] 5.1 Update docs/deployment.md with production deployment steps
- [x] 5.2 Add Traefik configuration guide
- [x] 5.3 Document subdomain scheme and DNS requirements
- [x] 5.4 Update README.md with deployment section
## 6. Testing & Verification
- [x] 6.1 Test label generation for all built-in tools
- [x] 6.2 Verify Traefik routes correctly in local stack
- [x] 6.3 Run backend tests: `cd apps/api && pytest`
- [x] 6.4 Run linters: `ruff check app/` and `mypy app/`
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-14
+65
View File
@@ -0,0 +1,65 @@
## Context
OpenCode is an AI-powered terminal-based development environment with a web interface. Unlike code-server which is a full web IDE, OpenCode provides a terminal experience accessible through the browser. This POC validates that the spawn system handles different runtime types including web terminal forwarding.
Current state:
- OpenCode manifest exists in apps/api/app/tools/manifests/opencode.yml
- No container image or runtime defined yet
- No health reporting mechanism
- Spawn infrastructure will be built in FN-010
## Goals / Non-Goals
**Goals:**
- Define OpenCode as a spawnable tool
- Provide web terminal interface in container
- Report health status (running/idle/error)
- Support interactive terminal sessions
**Non-Goals:**
- Full task queue or job scheduler
- Persistent process management
- Log streaming (deferred)
- Multi-language support beyond terminal
## Decisions
**1. Use official OpenCode Docker image**
- Rationale: Maintained, includes AI features and web terminal
- Alternative: Custom image - unnecessary for POC
**2. OpenCode runs as a persistent container**
- Rationale: Easier to manage lifecycle (start/stop/status). Terminal sessions need persistent container.
- Implementation: Container runs OpenCode with web interface on port 3000
**3. Health check via HTTP endpoint**
- Rationale: Standard Docker health check mechanism. Traefik can use it.
- Endpoint: `GET /` returns 200 when ready
**4. Workspace mounted from host (same as code-server)**
- Rationale: Consistency. Shared workspace between tools.
- Path: `/data/workspaces/{user_slug}/{project_slug}`
**5. Configs/secrets injected same as code-server**
- Rationale: Reuse FN-009 infrastructure. No special handling needed.
## Risks / Trade-offs
**[Risk] OpenCode container requires significant resources**
→ Mitigation: Set resource limits (4GB RAM, 2 CPU). Document requirements.
**[Risk] Web terminal performance over slow connections**
→ Mitigation: Use modern terminal emulation with compression. Document bandwidth requirements.
**[Risk] AI features require API keys**
→ Mitigation: Support secret injection for API keys. Document configuration.
## Migration Plan
No migration. New feature.
## Open Questions
1. Should OpenCode support multiple terminal sessions?
2. Do we pre-configure common development tools?
3. Should OpenCode integrate with the platform's AI provider?
@@ -0,0 +1,28 @@
## Why
OpenCode is an AI-powered terminal-based development environment that provides a web interface for interactive development. It demonstrates the platform's extensibility beyond standard tools like code-server. As a POC, it validates the manifest-driven spawn system with a non-trivial runtime that requires web terminal forwarding.
## What Changes
- **OpenCode manifest**: Define the tool with terminal web interface, workspace mounts, and health checks
- **Container image**: Reference to OpenCode image with built-in web terminal
- **Health reporting**: Endpoint that reports tool health to the platform
- **Spawn integration**: Reuse the spawn flow from FN-010 but with OpenCode-specific configuration
- **Web terminal**: Support for browser-based terminal access
## Capabilities
### New Capabilities
- `opencode-manifest`: OpenCode tool manifest with web terminal config
- `web-terminal`: Support for browser-based terminal interfaces
- `health-reporting`: Tool health status reporting mechanism
### Modified Capabilities
- None (reuses spawn infrastructure from FN-010)
## Impact
- **apps/api/app/tools/manifests/opencode.yml**: Updated manifest
- **apps/api/app/services/spawn.py**: Minor updates for OpenCode-specific mounts
- **apps/web/src/**: OpenCode appears in tool selection UI
- **Docker images**: Uses official OpenCode image
@@ -0,0 +1,27 @@
## ADDED Requirements
### Requirement: Container provides web terminal interface
The system SHALL provide a web terminal interface in the OpenCode container.
#### Scenario: Terminal available
- **WHEN** the OpenCode container is running
- **THEN** a web terminal is accessible via HTTP on port 3000
- **AND** the user can execute shell commands through the browser
#### Scenario: Workspace access
- **WHEN** the container runs
- **THEN** the project workspace is mounted at /workspace
- **AND** the user can read/write files in the workspace
### Requirement: Container supports AI features
The system SHALL allow AI-powered development features in the OpenCode environment.
#### Scenario: AI assistance
- **WHEN** the user interacts with OpenCode
- **THEN** AI features are available for code completion and assistance
- **AND** the user can configure AI provider settings
#### Scenario: Terminal session persistence
- **WHEN** the user opens a terminal session
- **THEN** the session persists while the container runs
- **AND** multiple sessions can be opened
@@ -0,0 +1,22 @@
## ADDED Requirements
### Requirement: Tool reports health status
The system SHALL provide a mechanism for OpenCode to report its health.
#### Scenario: Health endpoint
- **WHEN** the OpenCode container is running
- **THEN** it exposes a / endpoint for health checks
- **AND** returns 200 when the web terminal is ready
#### Scenario: Health check in Traefik
- **WHEN** the container is spawned
- **THEN** Traefik uses the health endpoint for routing decisions
- **AND** unhealthy containers are removed from the load balancer
### Requirement: Platform tracks tool health
The system SHALL track and display the health of OpenCode instances.
#### Scenario: Status display
- **WHEN** the user views an OpenCode instance
- **THEN** the current status is displayed (healthy, unhealthy, starting)
- **AND** the status updates automatically
@@ -0,0 +1,15 @@
## ADDED Requirements
### Requirement: OpenCode manifest defines web terminal environment
The system SHALL provide an OpenCode manifest with web terminal configuration.
#### Scenario: Manifest includes terminal config
- **WHEN** the OpenCode manifest is loaded
- **THEN** it specifies an OpenCode Docker image with web interface
- **AND** it defines exposed ports for the HTTP interface (port 3000)
- **AND** it defines volume mounts (workspace, config)
#### Scenario: Manifest includes health check
- **WHEN** the manifest is used for spawning
- **THEN** it defines a health check endpoint
- **AND** specifies health check interval and timeout
+42
View File
@@ -0,0 +1,42 @@
## 1. Manifest Definition
- [x] 1.1 Create apps/api/app/tools/manifests/opencode.yml with web terminal config
- [x] 1.2 Add Docker image (ghcr.io/opencode-ai/opencode:latest), ports (3000), volumes
- [x] 1.3 Add health check configuration to manifest
- [x] 1.4 Validate manifest against ToolManifest schema
## 2. Container Setup
- [ ] 2.1 Verify OpenCode image availability and configuration
- [ ] 2.2 Document web terminal access pattern
- [ ] 2.3 Configure environment variables for terminal support
- [ ] 2.4 Test container locally with docker run
- [ ] 2.5 Verify web terminal accessibility
## 3. Spawn Integration
- [ ] 3.1 Verify SpawnService (FN-010) can spawn OpenCode instances
- [ ] 3.2 Add OpenCode-specific volume mounts (config)
- [ ] 3.3 Test spawn via API endpoint
- [ ] 3.4 Verify Traefik routing to OpenCode container
## 4. Frontend Integration
- [ ] 4.1 Add OpenCode to tool selection dropdown
- [ ] 4.2 Display OpenCode-specific options in spawn form
- [ ] 4.3 Show OpenCode instance status in detail page
## 5. Testing & Verification
- [ ] 5.1 Test terminal availability in spawned container
- [ ] 5.2 Test web interface accessibility
- [ ] 5.3 Test health endpoint response
- [ ] 5.4 Verify workspace mount is accessible
- [ ] 5.5 Run full test suite: `make test`
- [ ] 5.6 Run linters: `make lint`
## 6. Documentation
- [ ] 6.1 Document OpenCode setup in docs/development.md
- [ ] 6.2 Add OpenCode usage guide
- [ ] 6.3 Document terminal configuration and AI features