The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a command-line interface tool named drupalorg to interact with Drupal.org data. It executes drupalorg issue:show to retrieve issue details and comments, and drupalorg issue:link to provide a URL to the user.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes external, untrusted content from Drupal.org issue comments.
Ingestion points: Untrusted data enters the agent context through the output of the drupalorg issue:show <nid> --with-comments command.
Boundary markers: The instructions do not include explicit delimiters or 'ignore' directives to prevent the agent from following instructions potentially embedded in issue comments.
Capability inventory: The skill is limited to drafting text and summarizing information. It explicitly states that the CLI tool is read-only and cannot write back to the platform directly, mitigating the risk of automated malicious actions.
Sanitization: No evidence of sanitization or filtering of the external comment data is present in the skill instructions.

drupalorg-issue-summary-update