debugging-wizard

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill's primary function involves the agent ingesting and analyzing untrusted external data.
  • Ingestion points: The skill explicitly instructs the agent to parse error messages, trace execution flow through stack traces, and correlate log entries as part of its core workflow in SKILL.md and references/systematic-debugging.md.
  • Boundary markers: The instructions do not specify the use of delimiters or boundary markers to isolate processed log data from the agent's instructions, which could allow malicious content within a log to influence the agent's behavior.
  • Capability inventory: The instructions guide the agent in the use of powerful CLI tools (e.g., git bisect, pdb, dlv, node --inspect). While these tools are not provided by the skill itself, the agent's use of them on untrusted data poses a surface for exploitation if an injection occurs.
  • Sanitization: No sanitization, escaping, or validation steps are recommended for the external logs or error messages before the agent processes them.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 04:40 PM
Security Audit — agent-trust-hub — debugging-wizard