skill-creation-guide
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill suggests using WebFetch to retrieve content from 'https://raw.githubusercontent.com/gmh5225/awesome-skills/refs/heads/main/README.md'. This URL belongs to an untrusted GitHub account.
- [PROMPT_INJECTION] (LOW): There is an indirect prompt injection surface as the agent is encouraged to ingest external data. 1. Ingestion points: README.md from gmh5225. 2. Boundary markers: Absent. 3. Capability inventory: Web-fetching and instructions processing. 4. Sanitization: Absent.
Audit Metadata