rw-research-router
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied research materials and external academic links, which could potentially contain malicious instructions intended to influence the agent's behavior.\n
- Ingestion points: Processes user-provided documents and data (mentioned in SKILL.md) and fetches information from official research guideline websites (listed in references/source-map.md).\n
- Boundary markers: Instructions lack explicit delimiters to isolate external content from the system prompt.\n
- Capability inventory: The skill uses basic reading and routing capabilities without requesting advanced system permissions.\n
- Sanitization: No specific input validation or sanitization routines are provided for processed data.\n- [EXTERNAL_DOWNLOADS]: The skill references several well-known academic and research standards organizations, including the EQUATOR Network, PRISMA Statement, Cochrane, and JBI. These are established services and their inclusion for research validation purposes is standard practice.\n- [COMMAND_EXECUTION]: The package contains Python scripts (scripts/self_check.py and tests/test_journal_submission_route.py) designed to verify the skill's integrity and routing logic during development. These scripts perform routine file system operations and do not exhibit malicious behavior.
Audit Metadata