figure-legend-writer
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external figure descriptions and images, creating a surface for indirect prompt injection. \n
- Ingestion points: Ingests user-supplied figure descriptions and image file paths in
SKILL.mdandscripts/main.py. \n - Boundary markers: Instructions do not utilize explicit delimiters or warnings to ignore instructions embedded within user data. \n
- Capability inventory: Capabilities are restricted to text generation and writing output to local files; no network or shell execution tools are used. \n
- Sanitization: No sanitization or filtering of external content is applied before interpolation into formatting templates.\n- [SAFE]: The Python script in
scripts/main.pyuses only standard libraries (dataclasses,enum,argparse,pathlib) and does not perform network operations or access sensitive system paths.\n- [SAFE]: External references to thegithub.com/aipoch/medical-research-skillsrepository are consistent with the skill's authorship and represent standard vendor resources.
Audit Metadata