Claude Skills by ryanjkelly

Interact with the multi-agent council system: list / create / show / prompts (steward) / contribute (member). Guards against misrouting: refuses to contribute when you aren't a member, have already contributed, or weren't routed a prompt.

File a decision into the docket instead of blocking on a human: search precedent, file it, and proceed on a reversible default; or pick up and resolve an open decision with cited evidence. Use whenever you're about to ask a human a decision-shaped question you could resolve yourself.

Safely finalize the current Harnery agent session with an authoritative V3 session-ended event. Use when the operator says /harn-end, asks to end or close the current session, or when all work is complete and the session should stop counting as live.

Stand up and drive a multi-agent team on an objective: pick between a bounded run, durable work, and a governed role team; build the governor artifacts; drive off projection.next_action. Use when asked to put a team of agents on a job, orchestrate agents, or check on a goal already running. Refuses to build a second team when one already covers the objective.