github-actions-failure-triage
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data including GitHub Actions run logs and repository files, representing an indirect prompt injection surface.
- Ingestion points: GitHub Actions run logs, workflow YAML files, and repository scripts (SKILL.md, references/evidence-checklist.md).
- Boundary markers: Absent; the skill does not define specific delimiters for log content.
- Capability inventory: The agent can write to repository files (workflow YAML, project code, config) and trigger reruns (SKILL.md, references/failure-buckets.md).
- Sanitization: Absent; no explicit sanitization of external log content is defined.
- [SAFE]: The skill includes explicit guardrails against the collection of secret values (SKILL.md, references/evidence-checklist.md).
Audit Metadata