nanocoai/nanoclaw / .claude/skills/slack-agent-flow
slack-agent-flow
FlaggedLet an existing Slack agent create new agents that arrive as their own Slack bots — provisioned app, operator DM, and a shared three-way room, hot-started without a host restart.
container/agent-runner/src/mcp-tools/rooms.instructions.md
View on GitHub## Shared rooms (`create_room`, `add_to_room`)
A room is one Slack group conversation shared by the user and N agents, with a canvas tab carrying the room contract (purpose, members). `mcp__nanoclaw__create_room({ name, purpose, agents, include_me? })` opens one; `mcp__nanoclaw__add_to_room({ room, agent })` grows one.
### The team pattern — one room, not N
When the user asks for a TEAM (several agents for one project), never let each `create_agent` open its own room — that yields N separate three-way rooms nobody wants:
1. Create each agent with `room: 'none'` (they still get their operator DM).
2. When all are live, call `create_room` ONCE with all their names and a short public `purpose`.
For a SINGLE new agent, plain `create_agent` (default `room: 'own'`) is right — don't follow up with `create_room`.
### How it works
- `agents` takes the same names you use with `send_message` — agents you created or can already message. Unknown names come back as an error note, nothing half-created.
- Both tools are fire-and-forget and may require admin approval; the outcome arrives as a system note. When the room is live, the note tells you the destination to post to and whom to tag — post a brief intro in your own voice (1-2 lines, no mechanics, no member lists).
- In rooms, agents engage when @-mentioned; everything else accumulates as ambient context. Tag an agent to bring it in.
### Growing a room
`add_to_room` works, but Slack group conversations never grow in place — the room MOVES to a new conversation (everyone re-wired automatically; the old conversation keeps working). Prefer creating rooms complete: if you know the team needs four agents, create all four first, then one `create_room`.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |