public-commons-media-search
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were identified. The skill is designed for public media retrieval using structured API calls.\n- [EXTERNAL_DOWNLOADS]: The skill provides installation steps for related components from the author's own repository (AgentPMT/agent-skills), which is a standard procedure for this platform's ecosystem.\n- [PROMPT_INJECTION]: The skill facilitates the ingestion of external content from Wikipedia and Wikimedia Commons, presenting a potential surface for indirect prompt injection.\n
- Ingestion points: Data is fetched from article titles, excerpts, and media metadata across various Wikimedia projects (SKILL.md).\n
- Boundary markers: Instructions do not specify delimiters or constraints for interpreting retrieved content.\n
- Capability inventory: The skill is limited to read-only tool calls; no file-writing or system execution capabilities are defined in this skill.\n
- Sanitization: External descriptions are provided to the agent without explicit sanitization or filtering instructions.
Audit Metadata