kimi-datasource
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a legitimate gateway to diverse data repositories (e.g., IMF, SEC, World Bank) via a standard Model Context Protocol (MCP) implementation.
- [EXTERNAL_DOWNLOADS]: Connects to the vendor's API (api.kimi.com) to retrieve information. These network operations are strictly limited to the skill's primary function and target Moonshot AI's own infrastructure.
- [SAFE]: Accesses authentication tokens from the local .kimi-code/credentials directory. This is a standard and expected behavior for maintaining session state in the Kimi Code environment.
- [SAFE]: The Node.js MCP server (bin/kimi-datasource.mjs) includes logic to validate and restrict file write operations to intended paths, preventing path traversal vulnerabilities.
- [PROMPT_INJECTION]: Instructions in SKILL.md define a clear multi-step logic for data retrieval. While the skill ingests external data (Category 8 surface), the risk is mitigated by the structured tool definitions and specific task orientation.
Audit Metadata