qjudge-github-workflow-owner
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automates development tasks by executing local Git and GitHub CLI (
gh) commands through shell scripts. These scripts (scripts/commit-changes.shandscripts/create-pr.sh) are used to manage code flow. The implementation uses standard Bash quoting and array expansion to handle user-supplied strings safely and prevent shell injection. - [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it interpolates user-supplied data into PR and commit metadata.
- Ingestion points: User-provided strings for commit messages, PR titles, and PR bodies passed as command-line arguments to the skill's scripts.
- Boundary markers: Content is interpolated directly into commands without explicit delimiters or warnings for the agent to ignore embedded instructions.
- Capability inventory: The skill has the ability to modify the repository (commit/push) and create PRs on GitHub.
- Sanitization: Standard shell quoting is applied, which prevents command injection into the shell but does not filter the semantic content of the strings for LLM-targeting instructions.
Audit Metadata