resolve-problem-report

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install companion tools using the command npx skills add crissmoldovan/agent-skills. These resources are provided by the skill's author to extend the workflow's capabilities.
  • [COMMAND_EXECUTION]: To achieve its primary purpose of resolving software bugs, the skill utilizes the Bash tool to execute local build scripts, test suites, and reproduction commands. These operations are governed by a human-in-the-loop 'act gate' policy for any irreversible actions.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process problem reports which are natural language inputs from external users. It proactively mitigates the risk of indirect injection by requiring that inputs be restated as falsifiable claims and by explicitly excluding the original investigation transcript (which contains the raw report) from the context passed to the implementation tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 03:33 PM
Security Audit — agent-trust-hub — resolve-problem-report