agentleFS
haliphax-aihaliphax-ai/skills/ritual/SKILL.md

ritual

Skill0 starsChanged 8 months agoUnlicenseView on GitHub

Facilitates an isolated meeting, or "ritual" between multiple agents or subagents in order to brainstorm, provide feedback, execute concurrent work related to the same task, reach consensus, etc.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. afittestide/asimi★ 16same name
    2. ritual-foundation/ritual-dapp-skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository4 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 1 is distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When multiple agents need to communicate in the same shared context to brainstorm, provide feedback, execute concurrent work related to the …

    line 18
  2. Very specific:

    If the facilitator is not the agent handling the request, then the agent handling the request should hand this process off to the …

    line 32
  3. Very specific:

    If no id was provided, the facilitator should generate one

    line 33
  4. Very specific:

    If members includes subagents, the facilitator should spawn any requested subagents using the specified mode (default run).

    line 35
  5. Very specific:

    If members includes live agents, the facilitator should message those agents' sessions with instructions to read this skill document and …

    line 36
  6. Very specific:

    When the ritual prompt has been received, members should begin the discussion requested by the ritual prompt, following instructions from …

    line 41
  7. Very specific:

    If the facilitator is included in members, they should participate in the ritual discussion as well as facilitate it

    line 42
  8. Very specific:

    If a stalemate has been reached or the conversation is carrying on too long, the facilitator should request guidance from the user

    line 47
  9. Very specific:

    When the conversation has concluded (either from consensus or at the user's request), the facilitator should provide a summary of the …

    line 48

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Ritual

Overview

Facilitates an isolated meeting, or "ritual" between multiple agents or subagents in order to brainstorm, provide feedback, execute concurrent work related to the same task, reach consensus, etc. The ritual discussion takes place simultaneously using available rooms extension tools.

Requirements

The rooms extension must be installed. See rooms.

When to Use

When multiple agents need to communicate in the same shared context to brainstorm, provide feedback, execute concurrent work related to the same task/project, reach consensus, etc.

Workflow

Step 1: Accept Input

  • id - The unique identifier of the room ID so that it can be isolated from other rituals. If no ID is provided, one should be chosen by the agent handling the request to initiate a ritual.
  • facilitator - The agent ID of the agent responsible for facilitating the ritual. This agent can be a member of the ritual as well, but that is not required. The facilitator steers the conversation to remain on topic and surfaces any necessary information or requests to the user. The facilitator should provide regular status updates as to the progress of the ritual, since the user will not be able to see the discussion taking place. If no facilitator is provided, the current agent should assume the role. If an agent ID is provided, a new subagent for that agent ID should be spun up specifically for the ritual. If provided with an existing session ID, connect the live session's agent to the ritual as the facilitator.
  • members - A list of agents or subagents that will be added to the ritual and expected to participate in the discussion. Existing session IDs can be provided to include current sessions' agents; otherwise, new subagents of the specified agent IDs will be spun up specifically for the ritual.
  • mode - The spawn mode for subagents created for the ritual. Defaults to run (one-shot, internal only — no Discord threads created). Set to session for multi-phase rituals that require human input between steps (requires thread: true on spawn, which creates Discord threads). Use run for simple, single-phase rituals. Use session when the facilitator needs to steer agents between phases via sessions_send or subagents(action=steer).
  • purpose - The reason for the ritual. This could be a topic to brainstorm, an issue to troubleshoot, a concept to research, a request for consensus on a decision, or any other topic that would require coordination between multiple entities.

Step 2: Setup the Room

  • If the facilitator is not the agent handling the request, then the agent handling the request should hand this process off to the facilitator and step back
  • If no id was provided, the facilitator should generate one
  • The facilitator should join a new room with the appropriate ID (this will create the room if it doesn't exist)
  • If members includes subagents, the facilitator should spawn any requested subagents using the specified mode (default run). If mode is session, pass thread: true on the spawn call. Include instructions to read this skill document and await the ritual prompt
  • If members includes live agents, the facilitator should message those agents' sessions with instructions to read this skill document and await the ritual prompt

Step 3: Perform the Ritual

  • Once all ritual members have announced themselves in the room, the facilitator should provide them with the ritual prompt
  • When the ritual prompt has been received, members should begin the discussion requested by the ritual prompt, following instructions from the facilitator when provided
  • If the facilitator is included in members, they should participate in the ritual discussion as well as facilitate it
  • The facilitator should help keep the conversation on track, settle disputes, and regularly surface information about the ritual's progress to the user

Step 4: Conclude the Ritual

  • If a stalemate has been reached or the conversation is carrying on too long, the facilitator should request guidance from the user
  • When the conversation has concluded (either from consensus or at the user's request), the facilitator should provide a summary of the ritual to the user and instruct all members to leave the room

Cost Consciousness

Rituals multiply token usage roughly in proportion to the number of participating agents. Each agent runs its own session with full context (system prompt, skill instructions, room history), and every round of discussion means every agent reads and responds.

As a rough guide:

SetupApproximate Token Usage
Single agent task~15-25k tokens
4-agent ritual, 2 rounds~80-100k tokens
4-agent ritual, 4 rounds~150-200k+ tokens

The cost scales with:

  • Number of agents — each one is a full session
  • Number of discussion rounds — every round means every agent processes and responds
  • Context size — agents with larger system prompts or longer room histories use more input tokens

For cost-sensitive situations, keep rituals focused: fewer agents, fewer rounds, and clear prompts that minimize back-and-forth. A well-scoped 2-agent ritual can be just as effective as a 5-agent free-for-all.

Tool Reference

Room Tools

These are the core communication tools for the ritual.

  • room_join — Join a named room (creates it if it doesn't exist). Used by the facilitator and all members during setup.
  • room_send — Broadcast a message to all other members in the room. Primary communication method during the ritual.
  • room_recv — Receive pending messages from the room. Members should poll this to read what others have said.
  • room_list — List rooms you belong to, or list members of a specific room. Useful for the facilitator to verify all members have joined.
  • room_leave — Leave the room. Used by all participants when the ritual concludes.

Session Management

These tools handle spawning, messaging, and managing the agents involved.

  • sessions_spawn — Spawn a new subagent for a given agent ID. Used when members or facilitator specifies an agent ID rather than an existing session.
  • sessions_send — Send a message to an existing session. Used when members or facilitator references a live session ID.
  • sessions_list — List active sessions. Useful for resolving session references or verifying agent availability.
  • subagents (list/steer/kill) — Manage subagents spawned for the ritual. The facilitator can steer participants back on track or kill subagents when the ritual concludes.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.