| FMRS | 58 / 100 · C | 79 / 100 · B |
|---|
| Reliability | 8 / 20 | 12 / 20 |
|---|
| Security and permissions | 12 / 20 | 18 / 20 |
|---|
| Maintenance | 14 / 20 | 16 / 20 |
|---|
| Documentation | 12 / 20 | 18 / 20 |
|---|
| Setup experience | 12 / 20 | 15 / 20 |
|---|
| Best for | - Quantitative analysts and investors with deep research needs in the Korean stock market (KOSPI/KOSDAQ).
- Developers needing structured financial data in Python environments.
- Those who want to integrate financial analysis capabilities into AI assistants like Claude or Codex via MCP.
| - 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 | - Those needing real-time stock price or high-frequency trading data (data is prebuilt, potentially with some latency).
- Those focused only on US small-cap companies or deep EDGAR data (EDGAR support is limited).
- Users seeking a lightweight, single-purpose data source (it's feature-rich but heavy).
| - 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 | - Network access to download prebuilt data from HuggingFace, and access to DART and SEC APIs (EDGAR).
- Local file system permissions for caching data (default in user directory).
- For MCP configuration, need to modify config files for Claude Desktop or CLI.
- Optional: DART API key (for raw data re-collection, not needed for regular use).
| - 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 | - External disclosure content (e.g., news, filings) may be wrapped as untrusted when serialized, but AI should use it only as analysis basis, not execute any instructions within.
- Prebuilt data may not be up-to-date; be aware of time lags.
- When using remote MCP endpoint, ensure stable internet connection.
- Installation issues could arise if Python path is not set properly, affecting MCP configuration.
| - 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 Desktop, Claude Code, Codex, Cursor | Claude Code, Codex, Gemini CLI, Cursor, VS Code, GitHub Copilot, Continue |
|---|
| Tools | 22 | 0 |
|---|