claude-code-cicd
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Indirect Prompt Injection] (SAFE): The skill has a data ingestion surface as it is designed to analyze external CI/CD configuration files provided by the user.\n
- Ingestion points: Evaluates .github/workflows/.yml, .gitlab-ci.yml, and various shell scripts (.sh).\n
- Boundary markers: None are explicitly defined in the prompt instructions to delimit user data from system instructions.\n
- Capability inventory: The skill is restricted to diagnostic feedback and providing configuration patterns; it does not execute code, write to the filesystem, or make network calls.\n
- Sanitization: Includes a dedicated reference file (validations.md) that uses regex-based detection to identify and flag unsafe configurations within the ingested files.\n- [Data Exposure] (SAFE): No hardcoded credentials or sensitive file paths were found. The skill actively implements checks to prevent the accidental exposure of ANTHROPIC_API_KEY and other secrets.\n- [Remote Code Execution] (SAFE): No patterns for downloading or executing remote scripts or installing untrusted packages were detected.\n- [Prompt Injection] (SAFE): The instructions are technically focused on integration and do not contain any patterns aimed at overriding safety filters or bypassing agent constraints.
Audit Metadata