pr-fixer
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and implement instructions from GitHub PR comments, which are untrusted external inputs. A malicious actor could embed commands or instructions in a PR comment to manipulate the agent's behavior during the fix/commit cycle.
- Ingestion points: The agent is instructed to "Read all comments first" and "Pick a comment, fix it" in SKILL.md.
- Boundary markers: Absent. The skill does not provide delimiters or specific instructions to treat the content of PR comments strictly as data rather than executable instructions.
- Capability inventory: The agent has the capability to modify source code (file-write), execute git commands (commit), and interact with the GitHub API (network operations for replies and issue creation).
- Sanitization: Absent. There is no requirement for the agent to validate or sanitize the content of the comments before implementing suggested changes.
Audit Metadata