skills/haoxuanlithuai/awesome_cognitive_and_neuroscience_skills/Paper-to-Skill Extractor/Gen Agent Trust Hub
Paper-to-Skill Extractor
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized network operations were detected. The skill operates within its stated purpose of academic information extraction.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface due to its requirement to process untrusted external data (PDF files).
- Ingestion points: Data enters the agent via PDF file paths or pasted text in Phase 1 (Paper Ingestion).
- Boundary markers: The skill uses a multi-phase interactive workflow (Ingestion, Scanning, Selection, Generation) which requires explicit user confirmation at each step, acting as a human-in-the-loop boundary.
- Capability inventory: The agent uses file reading tools (PDF Read) and file writing capabilities to save generated SKILL.md files.
- Sanitization: A mandatory 'Phase 5: Self-Verification' protocol is implemented to check generated content against source locations, specifically flagging hallucinations or mismatches for correction.
Audit Metadata