The Agent Skills Directory

[SAFE]: The skill implements safety logic to protect main and master branches from force pushes and skips detached HEAD states, adhering to development best practices.
[COMMAND_EXECUTION]: Local shell execution is restricted to repository discovery and git operations. The scripts utilize set -euo pipefail and proper variable quoting to ensure execution stability.
[PROMPT_INJECTION]: The skill processes untrusted metadata from local git repository structures.
Ingestion points: scripts/git-push.sh reads branch names and repository paths using git branch and find.
Boundary markers: No explicit delimiters are used for the repository metadata processed by the script.
Capability inventory: The skill can execute git push across any discovered repository.
Sanitization: Bash variable quoting is used to prevent argument injection during command execution.

git-push