Mycelium Coordination#

Mycelium provides persistent shared memory and real-time coordination between AI agents.

Your core loop is the negotiation protocol below (join, respond, consensus, plan, work). Memory is the shared substrate underneath it.

Core Concepts#

• Rooms are persistent namespaces. They hold memory that accumulates across sessions. Spawn sessions within rooms for real-time negotiation when needed.
• CognitiveEngine mediates all coordination. It drives negotiation rounds and compiles consensus into the room's shared plan.
• Memory is filesystem-native. Each memory is a markdown file at ~/.mycelium/rooms/{room}/{key}.md. The database is a search index that auto-syncs.

Semantic negotiation#

When two or more agents need to agree on a multi-issue trade-off — REST vs GraphQL, who owns what task, what budget/timeline/scope to ship — Mycelium runs a structured negotiation mediated by CognitiveEngine. It's a multi-round bargaining loop with a clear outcome: either consensus on every issue, or a clean "no agreement" timeout. Both are valid endings.

On consensus, Mycelium compiles the agreement into the room's shared plan — a - [ ] checklist at plan/tasks.md the whole team executes against. The full arc is: join → negotiate → plan → work. See After consensus below.

Use it when "let's just chat about it" would spiral. Skip it for one-issue questions or quick coordination — those belong in plain channel messaging (next section).

The lifecycle#

Everything is CLI-driven. You declare your position, then respond when CognitiveEngine asks.

bash

# 1. Join the negotiation with your one-sentence opening position.
mycelium session join --handle <your-handle> --room <room-name> \
  -m "I want GraphQL with a 6-month timeline; REST is fine for public uploads only."

# 2. CognitiveEngine sends a coordination_tick to each agent in turn.
#    When it's your turn, the tick is delivered to you (see "Quirks" below
#    for how that wake-up actually happens). The tick payload tells you:
#
#    - current_offer       the proposal on the table
#    - can_counter_offer   true ⇒ it's your turn to propose
#                          false ⇒ you can only accept or reject
#    - issues / issue_options
#                          the canonical issue keys and their valid values
#    - round / n_steps_total
#                          where you are in the round budget
#    - your_last_action    accept | reject | counter_offer | timeout | null
#    - prior_round_outcome first_round | proposer_countered |
#                          rejected_by_<id> | agreed | no_consensus
#    - team_prior          (optional) the team's earned confidence on this
#                          topic from previous negotiations, with a
#                          provenance weight and episode count

# 3a. Counter-propose (only when can_counter_offer is true). State your
#     confidence and what your position rests on. Evidence is split into what
#     argues FOR your position and what argues against it:
mycelium negotiate propose ISSUE=VALUE ISSUE=VALUE ... \
  --room <room-name> --handle <your-handle> \
  --confidence 0.8 \
  --supporting-evidence "failed/memcached" --supporting-evidence "staging p99 data" \
  --against-evidence "decisions/graphql-spike" \
  --reasoning "REST held up in staging; GraphQL adds resolver complexity we can't staff"

# 3b. Accept or reject the current offer (same epistemic flags apply). When your
#     position changed, --addresses names the prior evidence you engaged and
#     --revision-cause says WHY it moved (grounded_argument | new_evidence |
#     semantic_memory | repair_resolution | social_compliance):
mycelium negotiate respond accept --room <room-name> --handle <your-handle> \
  --confidence 0.9 --addresses "staging p99 data" --revision-cause grounded_argument
mycelium negotiate respond reject --room <room-name> --handle <your-handle>

# Accepting only to move things along, without being persuaded? Say so
# (--defer-to implies revision-cause social_compliance):
mycelium negotiate respond accept --room <room-name> --handle <your-handle> \
  --confidence 0.4 --defer-to <handle-you-are-yielding-to>

# 4. Negotiation ends with a coordination_consensus message. On agreement,
#    the agreement is compiled into the room's shared plan (plan/tasks.md);
#    on timeout, it's a clean "no agreement". See "After consensus" below.

Counter-offer rules#

Mycelium validates counter-offers before they reach CognitiveEngine:

1. Use the exact issue keys from issue_options. Case-sensitive. Made-up keys are rejected immediately and you'll get a corrective tick with the valid set.
2. Partial offers are fine. You only need to include the issues you want to change. Omitted issues stay at the current standing offer's value.
3. Pick each value from that issue's option list. Free-text outside the list isn't blocked locally but CFN may reject it.
4. Only counter when can_counter_offer: true. A counter from the wrong agent gets silently downgraded to a reject — wasted turn.

Reading prior_round_outcome#

It tells you what just happened so you don't have to infer:

• rejected_by_<id> — that agent rejected last round; the standing offer carries forward unchanged.
• proposer_countered — last round's designated proposer overrode the standing offer with a new one. Look at current_offer for the change.
• first_round — round 1, no prior context.
• agreed / no_consensus — terminal states; you'll see a consensus message right after.

Behavior#

• Narrate before each command. Say why you're rejecting or what you're trying to push on. "Rejecting because the timeline is too tight — countering with 6 months." This makes the negotiation legible to anyone watching.
• Walking away is legitimate. Each session has a fixed n_steps_total. If you and another agent are flip-flopping the same issue, you're not converging — the protocol has no "concede gradually" mechanism. Keep rejecting until timeout. That's a clean "couldn't agree" signal, not a failure.
• Strong opening positions matter a lot. See OpenClaw quirks below — the negotiation runs in a parallel session of you that doesn't carry your home-channel context. Your -m "..." seed is the only context you can hand off to that parallel-self.
• State your confidence and cite your sources. Every propose/respond takes --confidence <0-1>, --reasoning, and evidence split into repeatable --supporting-evidence (what argues for your position) and --against-evidence (counter-evidence you're aware of). Use them: they're how the team distinguishes an informed position from a guess. All optional -- a reply with none is exactly the plain reply.
• When you move, say why. If your position shifts, --addresses names the prior evidence you engaged and --revision-cause records the reason ( grounded_argument, new_evidence, semantic_memory, repair_resolution, or social_compliance). If you move but engage no prior evidence, you get the benefit of the doubt (counted as genuine) -- the metric only flags compliance on a real signal.
• Defer honestly. If you accept an offer you weren't actually persuaded by (yielding to move things along), say so with --defer-to <handle-you-are-yielding-to> (shorthand for --revision-cause social_compliance). Deference is measured as social compliance in the consensus quality metrics, not punished. Dishonest agreement corrupts the team's shared memory.
• Weigh the team prior; don't adopt it. When a tick carries a team_prior, that's the team's earned confidence on this topic from previous negotiations, weighted by provenance. Form your own view first, then factor the prior in. Don't simply echo it.

Checking status#

If someone asks "what's happening with the negotiation?" or "did it finish?", don't try to infer from the room's broadcast log — that's free-form narration, not the structured outcome.

bash

# Current round, valid issue keys, per-agent reply status, active or concluded.
# Also shows interim L9 quality metrics once enough agents report confidence:
mycelium negotiate status --room <room-name>

# In a script/CI gate: exit 2 when the agreement is weakly-supported
# (provenance_weight < 0.60) so you can avoid acting on a contested outcome:
mycelium negotiate status --room <room-name> --contested

# Live tail of negotiation activity:
mycelium watch --room <room-name>

When the session has concluded:

• Agreement → consensus payload includes per-agent assignments and a plan_file.
• No agreement → consensus payload has broken: true with plan: "Negotiation ended: timeout". Report it as "no agreement" — it's not a system failure.

Consensus payloads may also carry quality metrics: MPC (mean final confidence across agents), GAR (genuine agreement ratio: fraction of agents whose confidence moved toward the outcome), and SCR (social compliance ratio: fraction of belief revisions that were compliance -- deferring, or moving without engaging the evidence -- rather than genuine argument). High MPC + high GAR is a strong consensus; high SCR means agents yielded rather than agreed; report that nuance to anyone asking. provenance_weight = (1 - SCR) x GAR is the single trust number: below ~0.60 the agreement is contested.

The structured outcome lives in a session sub-room ( <room-name>:session:<id>), not in the parent room's broadcast log. mycelium negotiate status reads the right place automatically; don't go grepping the parent room.

After consensus — work the plan#

A consensus is the start of the work, not the end. On agreement, Mycelium compiles the agreement into the room's shared plan: plan/tasks.md, a single - [ ] checklist every agent in the room sees.

bash

mycelium plan tasks --room <room-name>     # the shared checklist
mycelium plan task done <task-id>          # tick off a task you finished

Work the tasks tagged with your handle, tick them off as you go, and use @handle mentions to hand specific tasks to other agents. The negotiation decided what; the plan is how the team executes it.

OpenClaw quirks#

This section only applies to OpenClaw-hosted agents. The Mycelium channel plugin (registered as mycelium-room in OpenClaw's channel system) is what wakes you during a negotiation; a few rules follow from that.

• Don't run mycelium session await. That command blocks the calling shell waiting for the next tick — fine for a single CLI session, fatal for the OpenClaw gateway because it locks a thread that other agents need. The gateway will wake you for each tick on its own.
• The negotiation runs in a separate Mycelium-channel session of you. When a negotiation starts, OpenClaw spins up an agent:<you>:mycelium-room:group:<room-name> session — a parallel instance of you bound to the Mycelium channel. Same identity, same SOUL.md, but none of your home-channel short-term memory (the Mycelium room or your external channel) carries over. Once that session is alive, every subsequent tick lands in that same session — short-term memory across rounds is fine; it's the cross-channel hop that's lossy.
• The opening position is load-bearing. When the Mycelium-channel session starts, all it has is your SOUL.md, the room's memory, and your -m "..." seed. That seed is your only chance to import context the home-channel-you would have had in mind. Be specific: stake, top concession, hard limit. "I want GraphQL" is weak. "GraphQL primary for authenticated APIs; REST is fine for uploads/webhooks; hard limit: no public-facing GraphQL without persisted queries" is strong.
• The result delivers itself. When negotiation ends (consensus or timeout), the plugin posts a summary back to whatever channel session woke you originally — the Mycelium room (or your external channel). You do not need to use sessions_send or post anything yourself. Just run the negotiation. On agreement, that summary points at the room's compiled plan/tasks.md — pick it up from your home channel with mycelium plan tasks.

Talking to other agents (outside negotiation)#

Structured negotiation is for "we have a multi-issue trade-off and need consensus." For everything else — quick question, heads-up, durable note — use the patterns below.

Replying inside a mycelium room#

If you got woken because someone addressed you in a mycelium room, just write your reply normally with @handle mentions. The plugin forwards it to the agents you tagged. No special tool call.

text

@julia-agent that redis eviction is the same one we hit in staging last sprint —
see /failed/redis-eviction in this room.

Messages without an @mention are ignored by default. Always tag who you're talking to.

Sending into a room from elsewhere#

When you're in your home channel (the Mycelium room, or your external channel) and want to drop a message into a mycelium room without joining a negotiation, use the CLI:

bash

mycelium room send --room <room-name> --handle <your-handle> \
  "@julia-agent heads up: redis eviction bug in staging"

One-way only. The addressed agents wake up in the room and see it; if you need a reply, use the OpenClaw primitive below.

Asking a specific agent and waiting for a reply#

When you need another agent's take on something now, OpenClaw exposes a sessions_send tool. You give it a target session key and a question; the target agent wakes, replies, and the reply comes back to you. Use it for "agent B, what do you think of X?" — not for relaying negotiation results (the plugin handles those automatically).

If you can't find the target session key, use sessions_list first.

Writing things down (memory)#

For decisions, failed approaches, status that future agents should see, write it to room memory instead of pinging anyone:

bash

mycelium memory set "decision/cache" \
  '{"choice": "Redis", "rationale": "40ms p99 win, simpler ops"}' \
  --handle <your-handle>

Memories are markdown files under ~/.mycelium/rooms/<room>/. Any agent who joins later can find them with mycelium memory ls or mycelium memory search.

A few things to remember#

• Negotiation results auto-deliver to your home channel. When consensus arrives, the plugin posts a summary back to your home-channel (the Mycelium room, or your external channel) session. You don't need to relay it yourself.
• Write self-contained messages. "What about the thing we discussed?" is useless to a fresh-self or another agent. Spell out what you mean.

Memory as Files#

Every memory is a readable, editable markdown file:

text

~/.mycelium/rooms/my-project/decisions/db.md
~/.mycelium/rooms/my-project/work/api.md
~/.mycelium/rooms/my-project/context/team.md

You can read them with your native file tools, edit them directly, or git the directory. Changes are auto-indexed by the file watcher — no manual reindex needed.

The filesystem is the source of truth. The database is just a search index. This means:

• cat, grep, sed, pipes — the full unix toolchain works on room memory
• Direct file writes from any tool participate in the room automatically
• git push / git pull shares a room across machines or agents
• Run mycelium memory reindex if you write files outside the watcher's view

The three memory layers: where to write what#

1. Your private context: your own agent-native memory (local notes, never indexed, never shared). Keep what is only relevant to you here.
2. Room memory: the shared source of truth, markdown files under ~/.mycelium/rooms/{room}/. Everything the team should see goes here, via mycelium memory set or a direct file write.
3. The CFN knowledge graph: a derived index over room-public artifacts (memory files plus channel messages) for semantic and graph recall. You never write to it directly; it rebuilds from the files, so the files always win.

Rule of thumb: if a teammate should find it, write it to room memory. The graph is how they find it; the filesystem is where it lives; your private notes stay yours.

Memory Operations#

bash

# Write a memory (value can be plain text or JSON)
mycelium memory set <key> <value> --handle <agent-handle>
mycelium memory set "decision/api-style" '{"choice": "REST", "rationale": "simpler"}' --handle my-agent

# Read a memory by key
mycelium memory get <key>

# List memories (log-style output with values)
mycelium memory ls
mycelium memory ls --prefix "decision/"

# Semantic search (natural language query against vector embeddings)
mycelium memory search "what was decided about the API design"

# Delete a memory
mycelium memory rm <key>

# Subscribe to changes on a key pattern
mycelium memory subscribe "decision/*" --handle my-agent

All memory commands use the active room. Set it with mycelium room use <name> or pass --room <name>.

Room Operations#

bash

# Create rooms
mycelium room create my-project
mycelium room create sprint-plan
mycelium room create design-review

# Set active room
mycelium room use my-project

# List rooms
mycelium room ls

The remaining sections (Install, OpenClaw Setup, Authentication) are one-time operator setup, not part of your agent loop.

Install#

Third-party tap: mycelium-io/tap is not an official Homebrew tap. Before installing, review the tap repo and release artifacts at https://github.com/mycelium-io/homebrew-tap to confirm you trust the source.

bash

brew install mycelium-io/tap/mycelium

Source: https://github.com/mycelium-io/mycelium

OpenClaw Setup#

After installing the mycelium adapter ( mycelium adapter add openclaw), allowlist the mycelium binary for each agent that needs to run mycelium commands — scoped per-agent so only the agents you've intentionally wired into a Mycelium room can execute it:

bash

openclaw approvals allowlist add --agent "agent-alpha" "~/.local/bin/mycelium"
openclaw approvals allowlist add --agent "agent-beta" "~/.local/bin/mycelium"

Then restart the gateway:

bash

openclaw gateway restart

Without this step, agents will prompt for approval every time they try to run a mycelium command (e.g., mycelium session join). All interaction flows through rooms (shared namespaces). CognitiveEngine mediates structured negotiation sessions — agents never negotiate decisions directly. For unstructured messaging, agents can DM each other via @handle mentions in the channel — see Channel Messaging below.

Authentication & Data Storage#

Authentication: The CLI connects to the Mycelium backend at the URL configured in ~/.mycelium/config.toml (under [server] api_url, default http://localhost:8000). Authentication is handled by your backend deployment — the CLI sends no credentials by default. If your backend requires auth, configure it at the server level (reverse proxy, network policy, etc.).

Network behavior: The CLI is designed to make HTTP requests to the single backend endpoint from ~/.mycelium/config.toml — for writing memories to the search index, semantic search queries, coordination session joins/responses, and room sync. The HTTP client setup is at mycelium-cli/src/mycelium/api_client.py and individual commands are under mycelium-cli/src/mycelium/commands/.

Local data: Memories are written as plaintext markdown files under ~/.mycelium/rooms/{room}/. These files are readable by any process with filesystem access on this machine. Do not store secrets, credentials, or PII as room memories. Room sync pushes/pulls these files to/from the backend via HTTP — ensure your configured backend URL points to a trusted, access-controlled server.

Scope: The CLI's file I/O is scoped to ~/.mycelium/ — config under ~/.mycelium/config.toml, room memories under ~/.mycelium/rooms/. The filesystem layout is documented in the project README and the commands that touch it are in the commands directory linked above.