root-cause-debug
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional, providing a methodology for root-cause analysis without any executable scripts, binary files, or remote dependencies.
- [DATA_EXPOSURE]: The instructions explicitly restrict the agent from dumping environment variables, credentials, or customer data, and emphasize logging only non-secret data.
- [INDIRECT_PROMPT_INJECTION]: The content includes a defensive clause specifying that sources and quoted inputs must be treated as data and not as instructions, which helps mitigate risks associated with processing untrusted external content.
Audit Metadata