The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes scripts/test-hook.sh, a utility designed to execute local shell scripts provided as arguments for testing purposes. This script uses bash -c to pipe test data into the target hook script.
[COMMAND_EXECUTION]: The examples/load-context.sh script executes filesystem checks (e.g., looking for package.json, Cargo.toml, pyproject.toml) to detect project types and persist configuration via the $CLAUDE_ENV_FILE environment variable.
[EXTERNAL_DOWNLOADS]: The documentation in references/advanced.md provides examples of using curl to send notifications to Slack or metrics to external systems. These are code snippets for reference and are not executed by the skill itself.
[DATA_EXFILTRATION]: Example patterns in the documentation demonstrate how to log session metadata or hook outcomes to external databases and audit logs for monitoring purposes.
[PROMPT_INJECTION]: The skill provides guidance on creating 'Prompt-Based Hooks'. It includes defensive prompt engineering examples that instruct the model to specifically look for and block malicious activities like path traversal, credential access, and destructive operations.

Hook Development