What is coord-mem CLAUDE.md?

A project instruction file describing coord-mem, a graph-based memory system for AI agents. It documents the code layout, commands, design rules, storage, command-line tool, MCP server, and session recording.

It helps agents understand coord-mem’s memory model, find relevant files, use its tools, and review changes against the project’s conventions.

It gives the agent the project’s rules and structure before it changes the memory system, helping preserve decisions and invariants.

coord-mem CLAUDE.md — Instructions file for coding agents by khalildh: A project instruction file describing coord-mem, a graph-based memory

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.

agentmods

npx agentmods add instructions/khalildh/coord-mem/claude-md

Clone the repo

git clone --depth 1 https://github.com/khalildh/coord-mem

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/khalildh/coord-mem/claude-md.svg)](https://agentmods.dev/instructions/khalildh/coord-mem/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/khalildh/coord-mem/claude-md"><img src="https://agentmods.dev/badge/instructions/khalildh/coord-mem/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 1,310 This file is loaded in full into every session.

When invoked 1,310 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

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.

Model	Per session	Once invoked
Fable 5	$0.01310	$0.01310
Opus 5	$0.00655	$0.00655
Sonnet 5	$0.00262	$0.00262
Haiku 4.5	$0.00131	$0.00131

Measured 3d ago against content hash b61499e95c0c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coord-mem CLAUDE.md 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 3d ago.

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.

CLAUDE.md · 92 lines

How it starts

The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.

coord-mem

A memory-palace graph for AI agent memory: append-only nodes rooted at an origin, multi-parent provenance recorded as context sets, one chosen home per node (walkable palace tree), one auto-assigned lattice coordinate per node (its stable address; moves only with rehome), brightness from decayed use + likes, adjacency derived from PMI-weighted co-occurrence. Read DESIGN.md before changing the model — every structural decision there was argued for deliberately, including the non-goals.

Layout

src/decay.js — decayed counters (all salience reads apply the half-life)
src/core.js — CoordMem: nodes, provenance, placement, windows, PMI, retrieval, rooms, mis-shelving
src/store.js — JSON file persistence (atomic write via tmp + rename)
src/index.js — public exports
bin/coord-mem.js — thin CLI over the library
bin/mcp-server.js — MCP server (hand-rolled stdio JSON-RPC, still zero-dep); registered in .mcp.json with the project palace at palace.json
bin/record-window-hook.js — Claude Code Stop hook (wired in .claude/settings.json): scans the session transcript for coord-mem MCP tool activity since its last run and records one context window over every memory touched. Per-session offsets live in .coord-mem-hook-state.json; it must never block or fail the session.
bin/palace-view.js + web/palace.html — starfield renderer: zero-dep HTTP server (npm run view, --file/--port) serving a self-contained canvas app. Brightness = glow, PMI = spring gravity, placement = faint threads, mis-shelving = red ring + dashed suggestion. Layout seeding is hash-based (deterministic — keep Math.random() out of the renderer too). Polls /data every 3s, so it live-updates as memories accumulate.
test/ — node --test suite (includes an MCP client test that spawns the server and speaks the protocol)

Commands

npm test — run the test suite (built-in runner, no deps)
node bin/coord-mem.js <cmd> — CLI; store path from --file <path> or COORD_MEM_PATH, defaults to ./coord-mem.json. Run with no args for usage.
node bin/mcp-server.js — start the MCP server (stdio); same store-path rules. Tools: memory_add / get / like / record_window / walk / retrieve / neighbors / rooms / misshelved / rehome.

Read the full file on GitHub · 92 lines

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.

3d ago First seen · 92 lines · 1,310 tokens per session scan A b61499e95c0c

Subscribe to this mod's changes

coord-mem CLAUDE.md is an instructions file published in the GitHub repository khalildh/coord-mem (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,310 tokens to every session, about $0.0066 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 2026-08-31.

Memory and context

cognirepo CLAUDE.md

Instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).

ashlesh-t/cognirepo · 2,241 tokens

aeon CLAUDE.md

Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).

aeonfun/aeon · 3,720 tokens

inkwell-memory CLAUDE.md

Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.

veronchenko/inkwell-memory · 2,378 tokens

RNR-Enhanced-Cognee AGENTS.md

AGENTS.md instructions for vincentspereira/RNR-Enhanced-Cognee, covering rnr enhanced cognee implementation for codex, critical requirements, 1. ascii-only output (no unicode encoding), 2. dynamic categories (no hardcoded categories) and 3. standard memory mcp interface.

vincentspereira/RNR-Enhanced-Cognee · 6,855 tokens

memory-mcp-1file AGENTS.md

Instructions for pomazanbohdan/memory-mcp-1file, covering ⛔ l0 invariants (never violate under any circumstance), detection heuristic, 🔀 phase transition routing (⛔ blocking), gate-0: phase identification & loading and 🧾 proof-of-load requirement (critical).

pomazanbohdan/memory-mcp-1file · 7,991 tokens

openexp CLAUDE.md

Instructions for anthroos/openexp, covering openexp — development instructions, memory protocol (mandatory), before starting any task, after completing a task and when the user shares context.

anthroos/openexp · 799 tokens