The Agent Skills Directory

[SAFE]: The skill manages integration with established platforms (GitHub, Jira, Confluence) for issue tracking and documentation in a transparent manner.
[SAFE]: Sensitive files such as .auth.json and Playwright storage-state.json are explicitly handled within a .gitignore management logic to prevent accidental credential leakage to version control.
[SAFE]: The network interaction logic in issue-media.mjs and issue-tracker.mjs includes checks to ensure Authorization headers are only sent to trusted domains.
[SAFE]: External command execution (via git, gh, curl, and playwright) is confined to the skill's stated purpose of issue management and evidence capture.
[SAFE]: A sophisticated phase-based contract system is used to maintain the integrity of the agent's state across multiple steps, reducing the risk of unexpected behavior during the issue lifecycle.

issue-end