coding-debug
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of error messages and stack traces which could contain hidden instructions intended to manipulate the agent's behavior during the debugging process.
- Ingestion points: The agent is instructed to "Read error messages carefully" and "Read stack traces completely" in the first step of the debugging process (SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the contents of error logs purely as data or to ignore embedded instructions.
- Capability inventory: The skill grants the agent the ability to modify project files (Step 6) and add instrumentation code (Step 4), which could be exploited if the agent follows malicious instructions embedded in a stack trace.
- Sanitization: The instructions lack requirements for the agent to sanitize or filter external data before it influences the debugging logic or file modifications.
Audit Metadata