The Agent Skills Directory

[SAFE]: The skill consists of instructional content focused on improving the accuracy and scope management of an AI agent's edits. It defines a set of quality control rules (e.g., 'smallest reading wins', 'what worked keeps working') and a reporting format for verification.
[SAFE]: No malicious patterns, such as prompt injection, data exfiltration, or unauthorized command execution, were found in the provided files.
[SAFE]: The skill includes safety-conscious instructions, such as advising the agent never to approve a code diff it hasn't read.

scope