saved-searches
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external data from X/Twitter (search results and trending topics), which creates a surface for indirect prompt injection attacks. Malicious content within posts could attempt to hijack the agent's context.\n
- Ingestion points: Processes search result pages and trending topic lists via
savedSearchManager.js.\n - Boundary markers: No specific boundary markers or 'ignore' instructions for external data are provided.\n
- Capability inventory: Includes capabilities to navigate URLs, list data, export JSON, and delete account settings (saved searches).\n
- Sanitization: No sanitization or validation of the ingested social media content is described.\n- [NO_CODE]: The implementation scripts
src/saveSearch.jsandsrc/savedSearchManager.jsare missing from the provided skill files. This analysis is restricted to the documentation provided inSKILL.md.
Audit Metadata