pipeline-security
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill includes a 'Prompt Injection Safety Notice' that proactively instructs the agent to ignore any malicious instructions (like 'ignore previous instructions' or 'DAN' style prompts) found within analyzed CI/CD configurations. This correctly treats untrusted input as data rather than directives.
- [DATA_EXFILTRATION]: The skill implements exfiltration mitigations by instructing the agent to never include sensitive values like credentials or API keys in its output and by limiting its configuration to read-only tools (Read, Grep, Glob) with no network access.
- [COMMAND_EXECUTION]: The skill explicitly prohibits the execution of code, commands, or scripts found within the analyzed pipeline configurations, reducing the risk of accidental or malicious command execution during the review process.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests untrusted CI/CD files (Ingestion points: SKILL.md Step 1), it provides clear boundary markers (Safety Notice), limits capabilities to read-only tools, and requires output validation against a schema, effectively managing the injection surface area.
Audit Metadata