diagnose
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as logs, network traces, and error messages, which provides a surface for indirect prompt injection attacks.\n
- Ingestion points: Processes external logs, HAR files, and bug reports (SKILL.md), and captures direct user input via a shell script (scripts/hitl-loop.template.sh).\n
- Boundary markers: The skill instructs the agent to redact secrets and auth headers, but does not specify strict delimiters for all untrusted input sources.\n
- Capability inventory: The skill uses shell commands, test runners, and network tools (curl) to execute diagnostic tasks.\n
- Sanitization: The skill relies on manual redaction instructions rather than programmatic sanitization of incoming data.\n- [DYNAMIC_EXECUTION]: The skill facilitates the creation and execution of temporary scripts and test harnesses to reproduce bugs.\n
- Evidence: The file
scripts/hitl-loop.template.shacts as a template for generating interactive reproduction scripts, and SKILL.md explicitly directs the agent to create throwaway harnesses and CLI invocations at runtime.
Audit Metadata