error-diagnostics-smart-debug
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data to perform error analysis, which could contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: The skill ingests data from the
$ARGUMENTSvariable in the Context section and at the end ofSKILL.md. It also references an external fileresources/implementation-playbook.mdfor implementation examples. - Boundary markers: The instructions lack explicit boundary markers or delimiters to separate user-provided debugging data from the skill's own instructions, and there are no directives to ignore instructions embedded within the ingested data.
- Capability inventory: The skill is configured to use a specialized
Task tool(subagent_type="debugger") and describes a workflow that includes code analysis, interaction with observability platforms (Sentry, DataDog, etc.), and fix generation. - Sanitization: No sanitization, validation, or filtering of the input data from
$ARGUMENTSis specified in the instructions.
Audit Metadata