coordination-memory

coordination-memory is a skill for Claude Code, Codex from yanqiw/comem. It costs 44 tokens per session (503 once invoked), scanned A, original, MIT.

A set of instructions for coordinating multiple AI agents through a shared memory service. It defines how agents claim work, record evidence, hand off tasks, and obtain final acceptance from an integrator.

In plain words
What is it for?
It is for multi-agent software projects that need assignments, team boards, reviews, snapshots, contracts, and controlled handoffs.
Why use it?
It prevents agents from duplicating work or treating unfinished work as accepted. The protocol also keeps shared changes tied to the correct revision and assignment.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions AGENTS.md.

Good fit It is for multi-agent software projects that need assignments, team boards, reviews, snapshots, contracts, and controlled handoffs.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/yanqiw/comem/claude
View source ↗ yanqiw/comem
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add yanqiw/comem --skill claude
Clone the repo
git clone --depth 1 https://github.com/yanqiw/comem

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for coordination-memory

README.md
[![agentmods](https://agentmods.dev/badge/skills/yanqiw/comem/claude/github.svg)](https://agentmods.dev/skills/yanqiw/comem/claude)
Your own site
<a href="https://agentmods.dev/skills/yanqiw/comem/claude"><img src="https://agentmods.dev/badge/skills/yanqiw/comem/claude/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for coordination-memory

Your own site · 80×15
<a href="https://agentmods.dev/skills/yanqiw/comem/claude"><img src="https://agentmods.dev/badge/skills/yanqiw/comem/claude.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 44 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 503 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00044 $0.00503
Opus 5 $0.00022 $0.00251
Sonnet 5 $0.00009 $0.00101
Haiku 4.5 $0.00004 $0.00050

Measured today against content hash 42ad65126a29, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-14, from the pricing page.

Security

Grade A, and why

coordination-memory scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

src/coordination_memory_mcp/templates/claude-skill.md · 40 lines

What it actually says

Coordination Memory

Follow the Coordination Memory protocol in AGENTS.md at the repository root (the section between the coordination-memory markers).

Key rules:

  • completed_gate_passed is not accepted truth; only an integrator accept_event makes work accepted.
  • Always claim_assignment before working, and pass the current base_revision on mutating calls.
  • Discover with list_workspaces, then use get_workspace_detail (workspace_id) for compact teams and assignment counts. Choose one team and pass its required team_id to get_team_board, list_pending_reviews, get_snapshot, list_contracts, and get_attention_board before high-volume reads; there is no pagination.
  • Use get_event_detail, get_assignment_detail, get_run_detail, or get_contract_detail only after choosing a specific ID from a compact read.
  • get_team_board returns only assignment_id, title, and status. get_team_board.status is an optional, one exact status filter; omitting it excludes accepted, rejected, cancelled, and superseded. Call get_assignment_detail for the full task contract.
  • Only these eleven Ack tools return a compact acknowledgement without a payload echo: append_event, submit_handoff, heartbeat_run, checkpoint_run, raise_attention, record_run_binding, request_intervention, respond_intervention, review_event, accept_event, and reject_event.
  • claim_assignment and actor/workspace/team/assignment/contract mutations retain their existing responses; keep using active_run_id from claim_assignment.
  • For loop-dispatched work, treat the assignment as the task contract: read metadata.context_refs, obey allowed_paths and acceptance_criteria, and use metadata.session_bind.target_actor_id as the intended worker binding before any run exists. Keep run communication metadata current with record_run_binding after the conversation starts.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today First seen · 40 lines · 44 tokens per session scan A 42ad65126a29

Subscribe to this mod's changes

coordination-memory is a skill published in the GitHub repository yanqiw/comem (1 stars, last pushed today), licensed MIT. It adds 44 tokens to every session and 503 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-13.