research-writing-skill
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
figure_synthesis_guide.mdfile contains instructions for the agent to execute thepdflatexcommand to compile TikZ figures. This is a standard typesetting operation required for the skill's intended purpose in academic writing. - [PROMPT_INJECTION]: As outlined in
SKILL.mdandbrainstorming_guide.md, the skill is designed to ingest and process untrusted external data, including user research notes and advisor feedback. This represents a surface for indirect prompt injection, though the risk is mitigated by the skill's specific focus on academic prose and lack of high-privilege system operations.
Audit Metadata