cy-fix-reviews

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes issue files containing external PR review comments, creating an indirect prompt injection surface.
  • Ingestion points: Untrusted data enters the agent context through issue files located in the .compozy/tasks/ directory.
  • Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded instructions in the review content.
  • Capability inventory: The skill possesses significant capabilities, including project-wide file modification and the execution of shell commands.
  • Sanitization: No sanitization or validation of the review file content is performed before processing.
  • [COMMAND_EXECUTION]: The workflow triggers the execution of the repository's real verification commands through cy-final-verify. This execution primitive is a standard development capability but could be exploited if malicious commands are injected into the review files or if the repository's build scripts are compromised.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 12:15 AM