The Agent Skills Directory

[SAFE]: The skill performs legitimate repository initialization and repair tasks without malicious intent.
[SAFE]: Implements proactive security measures in resolve-config.sh, such as scanning for sensitive keys (e.g., 'api_key', 'token', 'secret') in configuration files and failing if any are detected.
[SAFE]: The Python utility run-store.py uses defensive programming techniques, including os.O_NOFOLLOW and os.O_DIRECTORY to prevent symlink-based attacks, and enforces strict owner-only access (mode 0700/0600) to workspace data.
[SAFE]: Command execution via subprocess.run is limited to Git operations with validated path components and environment isolation.
[SAFE]: External integrations with services like Linear, GitHub, and Plane are conducted through official MCPs or host-authenticated CLIs in a narrow, read-only capacity as defined in the provider profiles.

woostack-init