Skip to main content
Glama

Mesh

mcp_m7_mesh

Coordinate multiple M7 instances by syncing shared state, files, tasks, and messages across machines or locally so distributed agents can collaborate and resume work.

Instructions

M7 Mesh — Multi-instance communication and coordination.

Connects M7 instances across machines via relay server, or locally via SQLite.

Remote Relay Actions (set M7_MESH_RELAY, M7_MESH_SECRET, M7_WORKSTATION):

  • sync: Full sync — push state, pull peers/changes/messages

  • push: Push shared doc to relay (path, content)

  • pull: Pull shared doc from relay (path)

  • notify: Notify peers about file change (path, summary)

  • changes: View file change feed from peers

  • ask: Ask a question to specific peer or all (to, question)

  • reply: Reply to a message (message_id, answer)

  • inbox: View unread messages

  • broadcast: Send note to all peers (message)

  • log_task: Register M7 task in shared index (task_id, title, status, summary)

  • tasks: View shared task index

  • push-file: Read local file and push to relay (path) — auto-notifies peer

  • pull-file: Pull file from relay and save to local disk (path)

  • list-files: List all shared files on relay

Local Mesh Actions (legacy):

  • init: Initialize mesh (auto-detects remote if configured)

  • status: Show mesh health

  • announce: Publish context

  • peers: List peers

  • request_context: Query peer context

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "action": "<action>"
    },
    "name": "mcp_m7_mesh"
  }
}

Input Schema

NameRequiredDescriptionDefault
toNoTarget workstation for ask/message (or "all")
pathNoFile path for push/pull/notify
limitNoNumber of recent events/tasks to show (default: 20)
scopeNoContext scope (e.g., file path)
sinceNoTime filter: "1h", "24h", timestamp
titleNoTask title
actionYesMesh action to perform
answerNoAnswer text for reply
moduleNoModule name
ttl_msNoLock TTL in milliseconds (default: 30000)
contentNoContext content (JSON string or plain text)
messageNoMessage text for broadcast
peer_idNoSpecific peer instance ID to query
purposeNoPurpose of the lock
summaryNoChange summary for notify
task_idNoM7 cycle/task ID
questionNoQuestion text for ask
resourceNoResource key to lock/unlock (e.g., file path)
message_idNoMessage ID for reply
descriptionNoIntent description
file_actionNoFile action for notify: created/modified/deleted
intent_typeNoType of intent to propose
lock_actionNoLock operation
server_roleNoServer role description (e.g., "auth backend on server-1")
task_statusNoTask status: in_progress/completed/blocked
context_typeNoType of context to publish
project_rootNoProject root path (default: the active project)
expires_in_msNoContext TTL in milliseconds
files_changedNoList of changed files

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv3.22.2
    • changedInput schema / properties / project_root / description
      Previous value: -"Project root path (default: cwd)"New value: +"Project root path (default: the active project)"
  2. First observedv3.18.1

TDQS

A3.5/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden. It does add real context: the required environment variables for relay mode, that push-file auto-notifies peers, that pull-file writes to local disk, and that sync both pushes and pulls. It omits anything about failure modes, whether actions are idempotent, what authorization the secret implies, or the fact that many of these actions mutate shared state visible to other workstations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is front-loaded with the one-line purpose, then two clearly headed sections with one bullet per action, so scanning for an action is fast despite the length. It stays dense because the tool genuinely has 22 actions, though a few lines could be tighter and the 'local mesh' section omits task_lock, propose_intent, and events that appear in the enum, creating a partial inventory.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 29-parameter, 22-action tool with no annotations and no output schema, the description covers input routing well but says nothing about what any action returns — inbox, tasks, changes, and list-files all produce results an agent must interpret. It also omits the actions present in the enum but missing from both action lists, so the description is genuinely helpful for invocation but incomplete for interpreting and verifying results.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description goes further by pairing parameters to actions — (path, content) for push, (to, question) for ask, (message_id, answer) for reply, (path, summary) for notify. With 29 params and 22 actions, that action-to-parameter mapping is the single most useful thing here and the schema alone does not express it.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific capability — multi-instance communication and coordination — and enumerates 22 concrete actions split into remote relay vs local mesh, which is far more than a restatement of 'Mesh'. It falls short of a 5 because it never distinguishes itself from close siblings like mcp_m7_hive_coordinator, mcp_m7_cross_session, or mcp_m7_task_transfer, so an agent facing those tools still has to guess which one owns cross-machine coordination.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It implies the routing rule via the env-var note ('set M7_MESH_RELAY, M7_MESH_SECRET, M7_WORKSTATION') and labels local mesh actions 'legacy', which nudges an agent toward the relay path. But it never says when to call this tool instead of sibling coordination tools, and there is no guidance on which action to pick for a given goal (e.g., ask vs broadcast vs notify).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.