debug-helper
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [SAFE]: No executable code, scripts, or sensitive operations were detected. The skill is composed entirely of markdown instructions and a text-based report template.
- [NO_CODE]: The skill does not include any scripts, binaries, or configuration files that execute logic, which minimizes the technical attack surface.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process untrusted user inputs, including code snippets and stack traces.
- Ingestion points: Instructions in
SKILL.md(Step 1) direct the agent to collect error messages, stack traces, and relevant code provided by the user. - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings for the ingested content are defined in the instructions.
- Capability inventory: The skill defines no tool access or subprocess execution capabilities; its only functional capability is generating text based on a template.
- Sanitization: No input validation, escaping, or sanitization steps are specified for the ingested user content.
Audit Metadata