$npx -y skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builderPlan and build MCP servers with agent-friendly tools, schemas, error handling, and evaluation. Use when creating or refactoring MCP integrations.
| 1 | # MCP Builder |
| 2 | |
| 3 | Build MCP servers around user workflows, not raw API endpoints. |
| 4 | |
| 5 | ## Workflow |
| 6 | |
| 7 | 1. Read the target API docs and identify the workflows an agent must complete end to end. |
| 8 | 2. Design a small tool surface with high-signal outputs, clear identifiers, and actionable errors. |
| 9 | 3. Implement shared infrastructure first: auth, request helpers, pagination, truncation, and formatting. |
| 10 | 4. Add tool schemas and docstrings that make correct usage obvious. |
| 11 | 5. Evaluate the server with realistic tasks before expanding scope. |
| 12 | |
| 13 | ## Principles |
| 14 | |
| 15 | - Prefer workflow tools over thin endpoint wrappers. |
| 16 | - Return concise, high-signal responses by default. |
| 17 | - Use human-readable identifiers whenever possible. |
| 18 | - Make error messages corrective: tell the agent what to try next. |
| 19 | - Design for limited context and large datasets. |
| 20 | |
| 21 | ## Resources |
| 22 | |
| 23 | - `references/mcp_best_practices.md` for design principles that apply to every server. |
| 24 | - `references/python_mcp_server.md` for Python implementation patterns. |
| 25 | - `references/node_mcp_server.md` for TypeScript implementation patterns. |
| 26 | - `scripts/connections.py` and `scripts/evaluation.py` as repo-local helpers. |
| 27 | |
| 28 | ## Deliverables |
| 29 | |
| 30 | - A concrete tool inventory tied to user workflows. |
| 31 | - Strict input/output schemas. |
| 32 | - Evaluation prompts or scripts that confirm the server is usable by an agent. |