The Agent Skills Directory

[COMMAND_EXECUTION]: The skill implements a modular hook orchestrator script in examples/modular-pre-commit/pre-commit that dynamically discovers and executes scripts within the hooks.d/ directory. This is the intended primary function for providing a flexible hook system.
[DATA_EXFILTRATION]: Reference documentation includes examples of using curl to transmit repository metadata and commit details to external services such as Slack, Jira, and Jenkins for notifications and build triggers. These represent legitimate automation workflows.
[COMMAND_EXECUTION]: The documentation provides examples of using ssh for remote command execution to automate deployments on external servers, which is a standard DevOps practice.
[SAFE]: The skill provides patterns for processing untrusted data from Git references and external APIs. 1. Ingestion points: Commit messages, branch names, and API payloads from third-party integrations like Jira. 2. Boundary markers: Not explicitly utilized in the provided examples. 3. Capability inventory: Use of curl for network requests and shell execution for automation tasks. 4. Sanitization: Examples demonstrate basic validation using regular expressions but do not include complex escaping for external content.

git-hooks