grant-specific-aims-writer
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied research descriptions to draft grant aims. It lacks explicit boundary markers for this input, creating a standard indirect prompt injection surface.\n
- Ingestion points: User-provided study ideas and scientific questions in Step 1 of
SKILL.md.\n - Boundary markers: None.\n
- Capability inventory: The
main.pyscript has file write capabilities.\n - Sanitization: None.\n- [INDIRECT_PROMPT_INJECTION]: The skill includes an evaluation report (
eval_report_grant-specific-aims-writer_result.json) that asserts its own security status. These claims are treated as data and do not influence the independent security assessment.\n- [COMMAND_EXECUTION]: Thescripts/main.pyutility performs file system writes to a path specified by the user via the--outputargument. While standard for a document generation tool, this capability could be used to overwrite files if the execution environment is not restricted.
Audit Metadata