wp-ci-qa
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data from external PR comments, which presents a surface for indirect prompt injection attacks.
- Ingestion points: The
scripts/fetch-pr-context.shscript fetches metadata and comments from GitHub PRs usinggh pr view --json comments. - Boundary markers: The script uses simple text-based headers (e.g.,
--- @author (date) ---) to separate comments, which provide minimal protection against instructions embedded within the comments. - Capability inventory: The skill has significant capabilities, including local shell execution (PHP, Yarn, Composer), file system modifications (git commits), and network operations via the GitHub CLI (posting comments and updating labels).
- Sanitization: There is no explicit sanitization or filtering of the comment content before it is presented to the agent.
- [DYNAMIC_EXECUTION]: The skill generates and executes temporary scripts to handle specific WordPress environment requirements.
- The skill instructs the agent to create a temporary PHP file (
/tmp/abspath.php) containing adefinestatement and then uses theauto_prepend_filePHP configuration to execute it when running local tools like PHPCS or PHPUnit. This is a controlled execution path using a fixed template to bypassABSPATHenvironment guards. - [COMMAND_EXECUTION]: The skill makes extensive use of local command-line utilities to perform its tasks.
- It utilizes
gitfor branch management and committing code,gh(GitHub CLI) for PR interaction, and project-specific tools likecomposer,php, andyarnfor building and testing. These operations are consistent with the skill's stated purpose.
Audit Metadata