6 tools from the Cairn MCP Server, categorised by risk level.
View the Cairn policy →cairn_domain Read a specific .cairn/domains/<name>.md file — Layer 2 domain design context. Contains: current design, trajectory, rejected paths (with re-evalua... cairn_query Search .cairn/history/ entries — Layer 3 raw decision events. Returns full history entries sorted chronologically by decision_date. Use this to loo... cairn_doctor Run 'cairn doctor --json' and return structured health check results. Use at session start or after writing history/domain files to verify .cairn/ ... cairn_match Match keywords from the user's request against domain file hooks — precise machine-level intent detection without AI inference. Call this with keyw... cairn_output Read .cairn/output.md — the Layer 1 global constraint file. Contains: stage (project phase), no-go (banned directions), hooks (domain keywords), st... The Cairn MCP server exposes 6 tools across 3 categories: Read, Write, Execute.
Use Intercept, the open-source MCP proxy. Write YAML rules for each tool — rate limits, argument validation, or deny rules — then run Intercept in front of the Cairn server.
Cairn tools are categorised as Read (2), Write (1), Execute (3). Each category has a recommended default policy.
Deterministic policy on every MCP tool call. Per-identity grants. Full audit log.