unslop-review
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of instructional text and formatting examples. It does not contain any scripts, binaries, obfuscated code, or network-bound operations. The triggers and rules are focused on natural language formatting and tone adjustment.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, potentially untrusted content (pull request diffs and comments) provided by the user.
- Ingestion points: The agent ingests PR content and code review feedback provided by the user for the purpose of rewriting.
- Boundary markers: The instructions do not specify explicit delimiters or boundary markers to isolate user-provided data from the system prompt.
- Capability inventory: The skill explicitly prohibits high-risk operations, stating: 'Comments only. No commits, no git push, no auto-approve, no linter runs.'
- Sanitization: No explicit sanitization or filtering of input content is described.
- Risk Assessment: While the skill processes external data that could contain injection attempts, the lack of dangerous capabilities (like file writing or network access) and the narrow output focus (stylized text) render the risk negligible.
Audit Metadata