The Plan Review MCP is a remote server that checks an AI coding agent's implementation plan against your codebase's real conventions before any code gets written. It runs as a Model Context Protocol server, so Claude Code, Cursor, Codex, and any MCP-capable agent can call it directly during planning, catching drift, duplication, and unsafe schema changes before they exist, not after a pull request opens.
Four skills are live today. full-review runs all nine plan-check dimensions and is the default — run it on every plan before writing code. The other three are narrower, cheaper follow-up checks for a specific question mid-task.
| Skill | When to use it | Cost |
|---|---|---|
| full-review | Review this PRD/spec against our codebase — the default. | $1.00 |
| architecture-fit | Does this plan fit how our codebase is built? | 30¢ |
| schema-review | Is this migration safe? | 30¢ |
| security-review | What's the threat surface of this plan? | 30¢ |
Six tools cover the full workflow: discovering which skills exist, reading a skill's full instructions, resolving project context, running a review, fetching a persisted result, and looking up a project by name.
| Tool | Purpose |
|---|---|
| list_skills | List available plan-review skills plus planned ones. |
| get_skill | Full markdown instructions for a named skill. |
| get_project_context | Stack profile, ADR titles, module directories, and the skill registry for a project. |
| vibecop_check | Run a skill against a submitted plan and extracted intents; returns verdict, findings, recommendations. |
| get_plan_review | Fetch a persisted plan review by id. |
| resolve_project | Look up projects by name; returns candidates when ambiguous, never silently picks one. |
Mint a token in VibeCop Settings under "Plan Review MCP · Agent Auth," then add the server to your agent's MCP config. The full installer has copy-paste setup for Claude Code, Codex, and Cursor.