scenario-report
Warn
Audited by Snyk on Aug 16, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). [SKILL.md] describes a workflow where the agent gathers user-provided bug evidence (e.g.,
diagnostics_runand verbatim errors), redacts it, then performs search/dedupe and posts to the GitHub public tracker; the only outsider-authored free text ingested at runtime would be from GitHub issues (search results and matched issue bodies) during dedupe, which is community-authored content.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata