Active Learning Agent

Before installing or enabling this skill, confirm the following: - Permissions: Verify exactly which Feishu scopes the feishu-openclaw-plugin will be granted (read group messages, read 1:1/private messages, read docs and meeting transcripts, access calendar events, create docs, send messages). Only grant the minimum scopes you are comfortable with. - Consent & privacy: Because the skill reads private 1:1 chats and meeting transcripts, ensure affected users have given consent and this complies with your org's privacy policy. - Autonomy & limits: The skill is allowed to act ("能做的就做了") and is scheduled every 2 hours. If you want tighter control, disable the cron trigger, reduce frequency, or require manual approval for actions that send messages or create docs. - Data retention & storage: It writes to memory/insights-journal.md and other memory files — decide where those files live, who can access them, and how long to retain them. - Testing & scope restriction: Try the skill in a limited environment (one test user or test groups) before broad rollout. Monitor its outputs and ensure it properly attributes sources (SKILL.md emphasizes preventing misattribution — verify this in practice). - Audit & transparency: Keep an audit trail of messages the skill sends and periodic reviews of what sources it read. If you cannot verify the skill's provenance (source/homepage discrepancy), prefer caution. Technical note: No external network endpoints or environment variables are present in the SKILL.md; data flows remain within Feishu and the skill's memory files, but the main risk is accidental or undesired sharing of sensitive internal content via posted messages. If those behaviors are acceptable and properly consented to, the skill appears internally coherent with its stated purpose.