The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard system utilities including gh, jq, git, sed, and grep to interact with the local development environment and GitHub's project management infrastructure. These tools are used appropriately for querying and updating project metadata.
[INDIRECT_PROMPT_INJECTION]: The skill features functions that ingest data from external sources (GitHub API responses via gh CLI). While this constitutes an attack surface for indirect prompt injection, the skill employs robust JSON parsing with jq and implements explicit logic for validating status transitions, which serves to minimize the risk of processing malicious data payloads.
[SAFE]: The skill does not contain any hardcoded credentials, remote code downloads from untrusted sources, or obfuscation techniques. All operations are focused on project state enforcement as described in the metadata.

project-board-enforcement