krt-ci-questor
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data such as CI job logs, test reports, and workflow annotations, which serves as a surface for indirect prompt injection.
- Ingestion points: Processes external log content and metadata fetched via
gh,glab, andcircleciCLI tools (SKILL.md, Step 2). - Boundary markers: The instructions do not define explicit delimiters for untrusted data, though the use of a structured report template helps contain output.
- Capability inventory: The skill is capable of executing provider CLI commands and suggesting local shell commands for reproduction (SKILL.md, Step 4).
- Sanitization: The skill includes strong guardrails that forbid printing tokens or secrets and advise the agent to quote only the minimum necessary log fragments (SKILL.md, Guardrails).
Audit Metadata