The Agent Skills Directory

[SAFE]: The skill processes local project files and user-provided PRDs to assist in task management and issue creation. No malicious patterns, hardcoded secrets, or unauthorized data exfiltration mechanisms were detected.
[COMMAND_EXECUTION]: The skill guides the agent to draft runnable commands (such as pnpm test or curl) to be used as acceptance criteria in issue trackers. These commands are intended for documentation and later use by other agents or CI/CD systems, not for immediate execution by the skill. The risk of generating unsafe commands is mitigated by Step 5, which requires the agent to present all drafts to the user for approval before publication.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it reads content from PRDs or external plans. However, the instructions focus on structured data extraction and include a mandatory human verification gate (Step 5) to ensure that the resulting issue titles, descriptions, and check commands align with user intent and safety requirements.

loop-issues