What is engram CLAUDE.md?

Repository guidance for Engram, a small software library that gives AI agents memory across separate sessions. It explains the project’s purpose, tools, architecture, and development limits.

Use it when changing Engram’s JavaScript, SQLite database, MCP server, REST API, command-line tool, dashboard, or tests.

It gives the coding agent the background needed to work consistently in an unfamiliar codebase. This reduces guesses about the technology, file structure, and rules.

How HBarefoot/engram configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to HBarefoot/engram. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/HBarefoot/engram/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/HBarefoot/engram

Made for: Claude Code.

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/hbarefoot/engram/claude-md/github.svg)](https://agentmods.dev/instructions/hbarefoot/engram/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/hbarefoot/engram/claude-md"><img src="https://agentmods.dev/badge/instructions/hbarefoot/engram/claude-md/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/hbarefoot/engram/claude-md"><img src="https://agentmods.dev/badge/instructions/hbarefoot/engram/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 6,531 This file is loaded in full into every session.

When invoked 6,531 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5.1	$0.06531	$0.06531
Opus 5.5	$0.02612	$0.02612
Sonnet 5.5	$0.01306	$0.01306
Haiku 4.5	$0.00653	$0.00653

Measured 2mo ago against content hash 0f7fa6568bb4, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

engram 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 2mo 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 · 371 lines

How it starts

The opening of the file, as written. The whole thing — 371 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

Engram is a lightweight, embeddable memory layer that gives AI agents persistent, cross-session memory. It's designed as "SQLite for agent state" - any agent framework can plug into it.

Core Concept: Engram is NOT a RAG system, vector database, or chatbot. It answers "what does this specific person need me to know right now, given everything I've learned about how they work?" - like a colleague who's worked with you for years.

Tech Stack

Runtime: Node.js 20+ (ESM modules)
Database: better-sqlite3 (synchronous, embedded, zero-config)
Embeddings: all-MiniLM-L6-v2 via @xenova/transformers (~23MB, CPU-only)
MCP Server: @modelcontextprotocol/sdk (primary interface)
REST API: Fastify
CLI: Commander.js
Dashboard: React 18 + Tailwind CSS 3 + Vite
Testing: Vitest

Critical Constraints:

Plain JavaScript only (no TypeScript in v1)
No Express (use Fastify)
No ORMs (raw SQL with better-sqlite3)
No cloud dependencies, Docker requirements, or Python dependencies
Must work fully offline

Architecture

Three-layer architecture:

INTERFACES: MCP Server (primary) | REST API | CLI | GUI
     ↓
CORE ENGINE: extract/ | memory/ | embed/
     ↓
STORAGE: ~/.engram/memory.db (SQLite) | config.json | models/

Core Components

memory/store.js - SQLite CRUD operations
memory/recall.js - Hybrid search (embedding similarity + FTS + recency)
memory/consolidate.js - Duplicate detection, decay, contradiction flagging
extract/rules.js - Zero-dependency rule-based fact extraction
extract/secrets.js - Secret/sensitive data detection (CRITICAL: never store API keys)
embed/index.js - Embedding generation + model management (lazy download)
memory/feedback.js - Per-memory helpful/unhelpful votes; derives feedback_score, may auto-adjust confidence
memory/context.js - Backs the engram_context MCP tool; markdown/xml/json/plain output with token budgeting
memory/health.js, memory/analytics.js - Back the dashboard Health/Statistics pages and /api/analytics/* endpoints
server/mcp.js - MCP server with 6 tools (remember, recall, forget, feedback, context, status)
server/rest.js - Fastify REST API + dashboard serving
import/ - Document import: wizard.js, index.js, and 8 parsers under parsers/ (cursorrules, claude, package, git, ssh, shell, obsidian, env)

Read the full file on GitHub · 371 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.

2mo ago First seen · 371 lines · 6,531 tokens per session scan A 0f7fa6568bb4

Subscribe to this mod's changes

engram CLAUDE.md is an instructions file published in the GitHub repository HBarefoot/engram (9 stars, last pushed 4mo ago), licensed MIT. It adds 6,531 tokens to every session, about $0.0261 per session on Opus 5.5. 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 ai-agents ai-memory claude developer-tools embeddings javascript

engram CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
engram CLAUDE.mdHBarefoot/engram	Instructions file	9	4mo ago	MIT	A	6,531	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

engram: Instructions file for Claude Code