systematic-debugging

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use standard development and diagnostic tools including git, pytest, pip, and sqlite3 to gather evidence and verify fixes.
  • [DATA_EXPOSURE]: Includes detailed procedures for macOS-specific troubleshooting, such as reading application logs and local SQLite databases within sandboxed containers (~/Library/Containers/) to identify data corruption or XPC service issues.
  • [INDIRECT_PROMPT_INJECTION]: As a debugging utility, the skill is designed to process external data like error messages, stack traces, and logs. This represents a known attack surface common to diagnostic tools, which is addressed through the skill's structured multi-phase protocol.
  • [PRIVILEGE_ESCALATION]: Mentions tccutil reset for resetting macOS privacy permissions, which is a legitimate platform command for resolving sandbox-related file access issues during development.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 01:03 PM
Security Audit — agent-trust-hub — systematic-debugging