article-prompts-to-skills
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content (articles and prompt packs), creating a surface where instructions in a source article could theoretically influence agent behavior. However, the skill explicitly mitigates this through a structured extraction process.\n
- Ingestion points: Source articles or tutorials provided as input in SKILL.md.\n
- Boundary markers: The workflow requires an 'Extraction Ledger' and 'Portable Contract' to isolate logic from source presentation.\n
- Capability inventory: File system writes (folder creation), git operations (status, diff, commit), and execution of local validation scripts (quick_validate.py).\n
- Sanitization: Instructions mandate manual review of git diffs and browser-based verification of generated demos before completion.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute standard development commands for validation and repository management.\n
- Evidence: Use of git status, git diff, and git commit, along with the execution of a Python validation script (quick_validate.py) as described in the validation phase of SKILL.md.
Audit Metadata