Skills
skills/getsentry/sentry/sentry-backend-bugs/Gen Agent Trust Hub

sentry-backend-bugs

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses Read, Grep, Glob, and Bash tools to analyze local code repositories for specific bug patterns. All instructions are focused on legitimate quality assurance and bug detection. \n- [SAFE]: The skill references official Sentry repositories and documentation, which are consistent with the author's identity and its stated purpose. \n- [PROMPT_INJECTION]: The skill processes untrusted code diffs and pull requests, which is a surface for indirect prompt injection. This behavior is essential to its function as an audit tool. \n
  • Ingestion points: Code provided by the user, Warden findings, or branch diffs (SKILL.md). \n
  • Boundary markers: Absent. \n
  • Capability inventory: Bash, Read, Grep, Glob (SKILL.md). \n
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 02:22 PM