reusable-workflow-patterns
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSNO_CODE
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill is designed to ingest and parse untrusted CI/CD configuration files (e.g., .gitlab-ci.yml, Jenkinsfile) from external repositories to identify patterns.
- Ingestion points: Repository files across supported CI/CD platforms as defined in the resolution workflow.
- Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the source CI/CD files during the parsing phase.
- Capability inventory: The agent uses mcp_github_search_repositories and file-writing capabilities to generate new workflow files. It does not directly execute code from the ingested files.
- Sanitization: Mitigation is present in the form of mandatory actionlint validation and strict requirements for pinned commit SHAs and verified publishers for all generated outputs.
- [EXTERNAL_DOWNLOADS]: The skill templates reference official GitHub Actions from the 'actions' organization, which is a trusted publisher. These references are documented neutrally as they are part of the intended workflow generation logic.
- [NO_CODE]: The skill consists exclusively of instructions, templates, and documentation standards without any included executable scripts or binaries.
Audit Metadata