The Agent Skills Directory

[SAFE]: The skill implements a robust security model for board configuration. It includes clear boundaries for operations, specifically excluding the creation or deletion of work items, issue transitions, and repository changes. A key security feature is the requirement for explicit confirmation before any configuration changes, copies, or normalizations are applied.
[SAFE]: The packaged GitHub helper script (setup-github-board.mjs) uses the official GitHub CLI (gh) for all operations. It implements dry-run behavior by default, requiring the --apply flag to execute mutations. The script also includes comprehensive guards to prevent creating duplicate 'Priority' fields when a native organization-level Priority field already exists.
[SAFE]: Data ingestion is handled cautiously; retrieved names and descriptions from boards are explicitly treated as untrusted data within the workflow. The skill also enforces pagination and verify host/owner identity to ensure complete and accurate audits before proposing any changes.
[SAFE]: The skill maintains provider boundaries, ensuring that configuration instructions for one provider (like GitHub) are not incorrectly applied to another (like Jira). The Jira implementation relies on existing authenticated connections or browser capabilities rather than managing credentials itself.

project-board