skills/mattpocock/skills/diagnose/Gen Agent Trust Hub

diagnose

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill guides the agent to execute shell commands and scripts, including git bisect run, curl, and a human-in-the-loop bash template, to reproduce and diagnose software issues.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it processes untrusted data and maintains powerful execution capabilities.
  • Ingestion points: SKILL.md specifies reading external logs, traces, and HAR files; scripts/hitl-loop.template.sh captures user responses.
  • Boundary markers: Absent; there are no instructions to delimit or ignore instructions within the ingested artifacts.
  • Capability inventory: SKILL.md describes executing shell commands (e.g., git bisect, curl), running tests, and executing local scripts.
  • Sanitization: Absent; the skill does not specify any sanitization or validation for external data before it influences agent logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 07:00 AM