The Agent Skills Directory

[SAFE]: The skill implementation follows a robust human-in-the-loop workflow where the agent is explicitly instructed to propose actions and wait for human confirmation rather than executing them autonomously. This design provides a strong control mechanism over all operations.
[SAFE]: The skill does not contain any obfuscated code, remote dependencies, or suspicious command execution patterns. Its use of the Bash tool is scoped to project management tasks such as file fingerprinting and status checks.
[SAFE]: No sensitive data exposure or exfiltration vectors were identified. The ledger file and hot files are managed locally, and no network operations to external or untrusted domains are requested.

meta-prompt