The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains a Node.js script (scripts/apply-config.mjs) that users are expected to run locally to apply configuration changes to their environment. The script uses standard Node.js built-in modules and performs explicit file system operations.
[CREDENTIALS_UNSAFE]: The script reads and writes to the Claude Code configuration file located at ~/.claude/settings.json and project-specific .claude/settings.json. This access is core to the skill's functionality for managing tool permissions and feature flags, and the script includes a backup mechanism to preserve existing settings.
[DATA_EXFILTRATION]: No network activity was detected in the provided source code. The documentation mentions a traffic proxy (proxy.mjs) for token measurement, but the script is not included in the skill package and is described as a local tool for monitoring communication with the official Anthropic API.

context-diet