diagnose

Pass

Audited by Gen Agent Trust Hub on Jul 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill implements a structured debugging and diagnosis workflow. It guides the agent through systematic steps to reproduce and fix software bugs based on established engineering best practices.\n- [COMMAND_EXECUTION]: The skill utilizes a shell script (scripts/hitl-loop.template.sh) for Human-In-The-Loop (HITL) debugging. The agent is instructed to modify and execute this script locally to gather feedback from the user. The script's operations are limited to standard bash input/output and variable assignment.\n- [PROMPT_INJECTION]: The HITL script captures untrusted data (error messages) from the user which is then processed by the agent. \n
  • Ingestion points: scripts/hitl-loop.template.sh via the capture function's read command. \n
  • Boundary markers: Absent; the input is read directly into shell variables. \n
  • Capability inventory: The skill allows the agent to execute shell commands, perform instrumentation, and modify codebase files. \n
  • Sanitization: Absent; the script captures and prints raw user input.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 4, 2026, 02:08 PM
Security Audit — agent-trust-hub — diagnose