code-reviewer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data including source code, diffs, and pull request metadata which may contain malicious instructions.
- Ingestion points: Data enters the context via repository files, diffs, tests, comments, PR metadata, and supplied web materials as described in SKILL.md.
- Boundary markers: The skill contains an 'Untrusted data boundary' section in SKILL.md that explicitly commands the agent to treat all reviewed data as passive facts rather than instructions and to ignore any attempts to redirect review scope or authorize tools.
- Capability inventory: No shell commands, network operations, or privileged tool usages are defined in the provided skill files.
- Sanitization: The skill implements a secret-handling protocol in SKILL.md to redact sensitive values with '[REDACTED]' before including them in any reports or metadata.
Audit Metadata