| FMRS | 62 / 100 · C | 79 / 100 · B |
|---|
| Reliability | 9 / 20 | 12 / 20 |
|---|
| Security and permissions | 13 / 20 | 18 / 20 |
|---|
| Maintenance | 13 / 20 | 16 / 20 |
|---|
| Documentation | 14 / 20 | 18 / 20 |
|---|
| Setup experience | 13 / 20 | 15 / 20 |
|---|
| Best for | - Developers whose primary assistant is Codex, Cursor, or another non-Claude client and who want an independent Claude opinion
- Read-only, auditable code-review flows (Read/Glob/Grep access only)
- Users who want to keep context in one session and ask follow-up questions
| - Engineering teams scaling AI coding agents while controlling spend
- Platform teams that need auditable, rollback-aware agent execution records
- Organizations using multiple agents (Claude, Codex, Gemini) under one governed flow
|
|---|
| Not for | - Scenarios where Claude must edit or write to the codebase
- Users who want one registration covering all providers (Codex, Grok, Gemini, and so on) — use the unified @ask-llm/mcp package from the same repository
- Users unwilling to install and authenticate the Claude Code CLI
- Teams needing official Anthropic support or enterprise guarantees
| - Developers who just want a simple code-completion helper without run governance
- Teams unwilling to accept local JSONL run records with HMAC signing
- Environments without a supported adapter (Claude/Codex/Gemini CLIs, direct provider, or verifier-only)
|
|---|
| Required permissions | - Runs locally over stdio and invokes the installed, authenticated Claude Code CLI as a subprocess
- Read-only workspace access limited to Read, Glob, and Grep
- Reads the configured environment variables (model alias, fallback model, timeout, log level)
- Sends prompts and relevant file contents to Claude
| - Read/write access to the configured repo root (writable scope restrictable via --allow-path/--deny-path)
- Ability to invoke the chosen agent CLI (Claude/Codex/Gemini, etc.) and its underlying model account
- Local disk writes for run records, receipts, and share outputs (share/run-receipt., etc.)
- Execution of user-configured verifier commands (e.g. npm test)
|
|---|
| Risks and side effects | - Prompts and file contents leave the machine and are processed by Claude, so sensitive code or secrets need care
- Automatically invokes the Claude Code CLI, which may consume subscription or API quota
- The default opus model falls back to sonnet when overloaded, so results may come from a different model than expected
- Third-party, unofficial project with no affiliation with or endorsement by Anthropic
- Version 0.0.1, so interfaces and behavior may change
- get-usage-stats is in-memory only; stats are lost when the process restarts
| - Verifier commands are executed; poorly chosen verifier commands carry security risk despite preflight policy checks
- Cost and token figures may be estimates (labeled with provenance) and should not be treated as settled accounting
- Receipt integrity must be verified before use as trustworthy evidence; missing rollback evidence is flagged as EVIDENCE_BOUNDARY
- Policy defaults come from martin.config.yaml and can be overridden by CLI flags; misconfiguration can loosen safety boundaries
|
|---|
| Supported clients | Claude Code, Codex, Cursor, Claude Desktop | Claude Code, Codex, Gemini CLI, Cursor, VS Code, GitHub Copilot, Continue |
|---|
| Tools | 3 | 0 |
|---|