chisle-audit
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data (git diffs, files, and repository contents). While this creates an attack surface for indirect prompt injection, the skill's strictly read-only nature and lack of external capabilities minimize the risk.
- Ingestion points:
SKILL.mddefines targets asgit diffoutput, specific file paths, or the entire repository tree. - Boundary markers: The instructions do not define explicit delimiters or "ignore" instructions to separate target content from the skill's logic.
- Capability inventory: No capabilities for network operations, file writing, or command execution are requested or present in the instructions.
- Sanitization: There is no evidence of content sanitization or validation of the files being audited.
Audit Metadata