honeybadger
Warn
Audited by Snyk on Apr 30, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.70). The SKILL.md shows the agent uses the Membrane CLI to connect to Honeybadger and run actions like "list-faults" and "get-fault" (i.e., fetching Honeybadger project/fault/check-in data), meaning the agent will ingest and act on third-party, user-generated error/monitoring content from the Honeybadger service which could contain attacker-controlled text and influence subsequent actions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata