Design GET /inventory/search endpoint
Support filtering by SKU, category, and stock status with pagination.
API Development Kanban Board Template for Engineering Teams
Ship well-designed, tested, and documented API endpoints with a Kanban board built specifically for backend and API teams. This template tracks each endpoint from initial backlog idea through schema design, implementation, testing, and final documentation before it ships to production. Built for use with the MCP server, an AI agent can automatically move tasks between columns as code is committed, tests pass, and docs are published—keeping your board accurate without manual updates. Whether you're building a small REST API or scaling a public API surface, this board gives your team clear visibility into what's designed, what's in progress, and what still needs docs.
Template preview
Endpoints and API changes that are identified but not yet designed
Design GET /inventory/search endpoint
Support filtering by SKU, category, and stock status with pagination.
Plan rate limiting for public API
Define request quotas per API key tier and 429 response format.
Scope webhook delivery retries
Determine backoff strategy for failed webhook deliveries.
API contract and schema design in progress, including request/response shapes and auth requirements
Draft schema for POST /orders
Define request body validation rules and response payload including order_id and status.
Design pagination for GET /customers
Decide between cursor-based and offset-based pagination, document query params.
Define error response format for 4xx/5xx
Standardize error object shape with code, message, and field-level details.
Endpoints currently being implemented against the approved design
Implement PUT /users/{id}/profile
Handle partial updates, validate email uniqueness, return updated resource.
Implement DELETE /sessions/{token}
Invalidate session token and clear associated cache entries.
Add auth middleware for /admin/* routes
Restrict access to admin-role tokens, return 403 on failure.
Endpoints with implementation complete, undergoing unit, integration, or manual QA testing
Write integration tests for POST /orders
Cover happy path, invalid payload, and insufficient inventory scenarios.
Load test GET /inventory/search
Verify response times under 500ms at 200 requests/sec.
Endpoints tested and passing, now being documented for the public API reference
Document POST /users/login in API reference
Add request/response examples, auth requirements, and error codes.
Update REST API changelog for v2.3
Summarize new endpoints and breaking changes for consumers.
Endpoints fully implemented, tested, documented, and merged to main
GET /health endpoint shipped
Basic liveness check endpoint deployed and documented.
POST /auth/refresh shipped
Token refresh flow implemented, tested, and documented.
Step 1
Add new endpoint ideas or required API changes to Backlog as they're identified.
Step 2
Move tasks to Design once the team is ready to define request/response schemas and auth rules.
Step 3
Shift tasks to In Progress when a developer starts implementation against the finalized design.
Step 4
Move completed implementations to Testing for integration/load testing, then to Documentation once tests pass.
Step 5
Mark tasks Done after documentation is published and the endpoint is merged to main.
AI agent usage
This playbook enables an AI coding agent to manage an API development board through the MCP server, moving endpoints through design, implementation, testing, and documentation stages automatically as work progresses. The agent uses list_tasks and get_task to check current state, then update_task or move_task to reflect progress, keeping the kanban board synced with real development activity without manual upkeep.
## API Development Board (MCP)
This project uses a Kanban board with columns: `Backlog`, `Design`, `In Progress`, `Testing`, `Documentation`, `Done`.
When implementing an endpoint:
1. Use `list_tasks` to find the relevant task in `Design` or `Backlog`.
2. Use `move_task` to shift it to `In Progress` when coding starts.
3. Once unit tests pass, `move_task` to `Testing`.
4. After docs are written, `move_task` to `Documentation`, then `Done` on merge.
5. Use `update_task` to append notes (e.g., endpoint path, status codes) to the task description.
Always check `get_board` before creating a new task with `create_task` to avoid duplicates.Learn more about the board-centric workflow in Kanban for AI Agents or open the MCP guide.
The free plan supports 1 project with up to 25 tasks total. This API Development Board template uses 15 sample tasks across 6 columns, leaving room to add more within the free tier.
Yes, you can customize columns using the kanban board interface. This template ships with 6 columns (Backlog, Design, In Progress, Testing, Documentation, Done) but you can adjust to fit your team's workflow.
The MCP server exposes tools like list_tasks, move_task, and update_task so an AI agent can query task status and move endpoints through columns automatically as code is written, tested, and documented.
You'll need to upgrade to the pro plan ($12/mo or $96/yr) to create additional tasks beyond the 25-task limit, or archive/delete completed tasks to stay within the free tier.
Yes, the macOS app supports full kanban board functionality including columns and tasks, and syncs with the same REST API and MCP server used for automation.
Yes, keyboard shortcuts and the command palette let you quickly create, move, and search tasks without leaving the keyboard, speeding up board management during active sprints.
Ready to create this board?
Sign in and use the template to create a project with columns and sample tasks.