Human In The Loop
Decide when work needs a human decision and route it as a durable, actionable queue item.
- Repository
- mvschwarz/openrig
- Stars
- 8,553
- Licence
- Apache-2.0
- Files
- 1 file
What it does
This skill teaches an AI agent to tell apart three situations at the end of a work slice: finish cleanly and continue automatically, stop and ask a person for a real decision, or deliberately park the work. When a human is needed, the agent is guided to write a clear request with proof, the decision to make, a recommended default, and the possible outcomes, rather than dropping a loose chat message. It also covers designing a dashboard that puts the decision queue first.
Great for
- Developers building or operating multi-agent systems in the OpenRig daemon who need clear human approval gates
- Teams designing human review queues or Mission Control style dashboards for agent work
- Anyone who wants agents to avoid pestering humans with routine status updates
Skip it if
- Simple one-off chat tasks, since the skill assumes a daemon-backed rig with queues and workflows (the referenced Mission Control and queue files are not included here)
- Handing work between agents, where the SKILL.md says to use queue-handoff instead
- Teams expecting an approval step after every task, because the skill explicitly says not to create human gates for clean closeouts
Try asking
- “Classify the closeout of this slice: auto-continue, human gate, or park?”
- “Write a human queue item asking for a decision on the provider auth failure, with a recommended default”
- “Design the dashboard so the pending decision queue is the primary view and rig state is secondary”
- “Review whether this completed slice actually needs human approval or can continue to the next named slice”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.
human-in-the-loop.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip human-in-the-loop.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip human-in-the-loop.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give the agent a PROGRESS.md or workstream plan that names the next safe slice, so it can judge clean closeouts correctly.
- Ask for the recommended default and action outcomes explicitly so the human gets a decision-ready request.
- Check the referenced sibling skills (queue-handoff, watchdog, attention-queue) if you want the full system; this skill alone is mostly guidance.
What's inside
- SKILL.md
Redistributed unmodified from mvschwarz/openrig under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.