Skills
skills/ibreez3/xiaohongshu-skill/search-feeds-skill/Gen Agent Trust Hub

search-feeds-skill

Pass

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill ingests untrusted data from external Xiaohongshu search results.\n
  • Ingestion points: scripts/search-feeds.mjs retrieves results from the XIAOHONGSHU_MCP_URL endpoint.\n
  • Boundary markers: The script does not use delimiters or boundary markers to distinguish external content from the agent's context.\n
  • Capability inventory: The skill uses Bash to run scripts and performs network requests via fetch.\n
  • Sanitization: No sanitization or validation is applied to the content retrieved from the search service.\n- [DATA_EXFILTRATION]: The skill performs network operations using the fetch API. It targets an endpoint specified by the XIAOHONGSHU_MCP_URL environment variable, which defaults to localhost.\n- [COMMAND_EXECUTION]: The skill executes its search logic by invoking a Node.js script through the Bash tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 8, 2026, 08:25 AM