deliver
Fail
Audited by Snyk on Nov 16, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E004: Prompt injection detected in skill instructions.
- Potential prompt injection detected (medium risk: 0.30). The skill is an orchestration workflow, and its only sketchy element is that it writes exact-path workspace-trust entries into the user's global Claude and Codex config files to get past harness trust dialogs, which weakens a protection beyond what the task strictly needs without malicious intent.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). Intake's required step reads a GitHub issue with its comments (
gh issue view N --comments), and anyone can comment on a public issue, so outsider text enters the run without the agent selecting a specific item.
Issues (2)
E004
CRITICALPrompt injection detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata