gh-issue
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill interacts with the official GitHub CLI (
gh) to perform repository management tasks. It utilizes secure shell scripting patterns, specifically quoted heredocs ('BODY'), to ensure that the user-provided issue content is treated as literal data and not executed as code by the underlying shell. - [COMMAND_EXECUTION]: The skill enforces a human-in-the-loop workflow, requiring the AI agent to present the final draft of the issue to the user for verification and potential modification before any commands are executed.
- [PROMPT_INJECTION]: While the skill ingests external data (user input for issue fields), the risk of indirect prompt injection is addressed through the use of clear markdown templates and the mandatory human review step before tool invocation.
Audit Metadata