principles-disciple@1.215.0
Stop correcting the same AI behavior across sessions. Turn repeated Agent corrections into Owner-approved, observable, reversible behavior principles.
AI security review
scanned · by lpm-firewall-aiThe bundle is a first-party OpenClaw extension that writes its own workspace guidance, including AGENTS.md and TOOLS.md, when the host builds a prompt and the files are missing. There is no npm install hook and no separate exfiltration or remote-payload chain, so the release warrants a warning rather than a block.