The Agent Skills Directory

[SAFE]: The skill serves as a text-based routing and teaching utility for an AI-native development loop. It analyzes project state via local configuration files (e.g., .and/config.yml) and workflow metadata to recommend the next appropriate skill or action.
[SAFE]: No malicious patterns such as remote code execution, data exfiltration, or credential harvesting were detected. The skill instructions explicitly prohibit the agent from automatically executing the next tool in the chain or modifying backend state.
[SAFE]: The skill uses disable-model-invocation: true in its configuration, which prevents the model from calling external tools during its execution context, further reducing the attack surface.

ask-andie