The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard local CLI tools including jq, git, and grep for project management tasks. These tools are used to parse the feature_list.json ledger and audit diffs to ensure that only the 'passes' boolean is modified during a session. This implementation follows best practices for local state verification.
[SAFE]: No malicious patterns such as prompt injection, Base64 obfuscation, or unauthorized data exfiltration were detected. The skill's architecture focuses on increasing the reliability and transparency of AI agent workflows through structured data and explicit verification steps.

feature-list-json