generate-ci-fix
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes sudo and chmod commands, which are standard operations for environment configuration and disk space management within CI/CD pipelines.
- [PROMPT_INJECTION]: The skill processes untrusted external data (CI logs and URLs), creating a surface for indirect prompt injection. Ingestion points: CI log files, pipeline URLs, and failure descriptions in the workflow parse input step. Boundary markers: None identified. Capability inventory: The skill can modify configuration files and execute validation commands. Sanitization: Not explicitly documented for log or URL processing.
- [EXTERNAL_DOWNLOADS]: The skill references trusted or well-known GitHub Actions such as shivammathur/setup-php and actions/cache for environment setup and caching.
Audit Metadata