drupal-work-on-issue
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple command-line operations including git commands (remote, diff, checkout, add, commit, push), PHP CodeSniffer (vendor/bin/phpcs), and the drupalorg CLI tool suite for managing issues and merge requests.- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon untrusted data from an external platform.
- Ingestion points: Untrusted content is ingested from the Drupal.org issue tracker and GitLab instance via
drupalorg issue:show,drupalorg mr:diff, anddrupalorg mr:logs. - Boundary markers: Absent. The skill does not utilize delimiters or specific instructions to isolate external data from its own execution logic.
- Capability inventory: The agent has significant capabilities including file system modification, version control operations (including pushing code to remote branches), and execution of the drupalorg toolset.
- Sanitization: Absent. Data retrieved from external sources is processed directly without filtering or verification of the content's safety.
Audit Metadata