Engrm

Cross-device memory for AI coding agents. Every session remembers what you learned yesterday — on any machine, for every team member.

npx engrm init

What It Does

Your AI agent forgets everything between sessions. Engrm fixes that.

• Captures automatically — hooks into Claude Code to record discoveries, bugfixes, decisions, and patterns as you work
• Remembers across devices — fix a bug on your laptop, continue on your desktop with full context
• Shares with your team — one developer's hard-won insight becomes everyone's knowledge
• Works offline — local SQLite is the source of truth; syncs when connected
• Guards your code — Sentinel audits changes in real-time before they land

Quick Start

1. Sign up

Visit engrm.dev and create an account.

2. Install

npx engrm init

This opens your browser for authentication, writes config to ~/.engrm/, and registers the MCP server + hooks in Claude Code. Takes about 30 seconds.

Alternative methods:

# From a provisioning token (shown on engrm.dev after signup)
npx engrm init --token=cmt_your_token

# Self-hosted Candengo Vector
npx engrm init --url=https://vector.internal.company.com

# Manual setup (air-gapped environments)
npx engrm init --manual

3. Use Claude Code normally

That's it. Engrm works in the background:

• Session start — injects relevant project memory into context
• While you work — captures observations from tool use (edits, writes, commands)
• Session end — generates a session digest, syncs to cloud, shows summary

━━━ Engrm Session Summary ━━━

📋 Request: Fix the OAuth redirect validation
🔍 Investigated: redirect_uri handling in auth.py
💡 Learned: scheme + host + port must all match registered URIs
✅ Completed: Stricter redirect_uri validation (auth.py)

🟢 Risk: Low (0.12)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

4. Check status

npx engrm status

Engrm Status

  User:        david
  Email:       david@example.com
  Device:      macbook-a1b2c3d4
  Plan:        Pro ($15/mo)
  Candengo:    https://www.candengo.com
  MCP server:  registered
  Hooks:       registered (6 hooks)

  Observations:    1,247 active
  By type:         change: 412, discovery: 289, bugfix: 187, ...
  Sentinel:        enabled (advisory, 142/200 today)
  Sync:            push 2m ago, pull 5m ago
  Security:        3 findings (1 high, 2 medium)

How It Works

Claude Code session
  │
  ├─ SessionStart hook ──→ inject relevant memory into context
  │
  ├─ PreToolUse hook ────→ Sentinel audits Edit/Write (optional)
  │
  ├─ PostToolUse hook ───→ extract observations from tool results
  │
  ├─ PreCompact hook ────→ re-inject memory before context compression
  │
  ├─ ElicitationResult ──→ capture MCP form submissions
  │
  └─ Stop hook ──────────→ session digest + sync + summary
        │
        ▼
  Local SQLite (FTS5 + sqlite-vec)
        │
        ▼ (sync every 30s)
  Candengo Vector (cloud)
        │
        ▼
  Available on all your devices + team members

MCP Tools

The MCP server exposes tools that Claude can call directly:

Observation Types

Features

Hybrid Search

Local FTS5 + sqlite-vec (all-MiniLM-L6-v2, 384 dims) combined with Candengo Vector's BGE-M3 semantic search. Results merged via Reciprocal Rank Fusion.

Sentinel — Real-Time Code Audit

LLM-powered review of every Edit/Write before it executes. Catches security issues, anti-patterns, and drift from team decisions.

⚠️  Sentinel: SQL query uses string concatenation instead of parameterized query
   Rule: sql-injection
   (Advisory mode — change allowed)

5 built-in rule packs: security, auth, api, react, database.

npx engrm sentinel init-rules        # Install all rule packs
npx engrm sentinel rules             # List available packs

Starter Packs

Pre-loaded knowledge for your tech stack. Detected automatically on session start.

Available: typescript-patterns, nextjs-patterns, node-security, python-django, react-gotchas, api-best-practices, web-security

npx engrm install-pack typescript-patterns

Secret Scrubbing

Multi-layer regex scanning for API keys, passwords, tokens, and credentials. Sensitive content is redacted before storage and sync. Custom patterns configurable in ~/.engrm/settings.json.

Observation Lifecycle

Observations age gracefully: active (30 days, full weight) → aging (0.7x search weight) → archived (compacted into digests) → purged (after 12 months). Pinned observations never age.

Pricing

Sign up at engrm.dev.

Self-Hosting

Point Engrm at your own Candengo Vector instance:

npx engrm init --url=https://vector.internal.company.com --token=cmt_...

Candengo Vector provides the backend: BGE-M3 hybrid search, multi-tenant namespaces, and team sync. See the Candengo docs for deployment.

Configuration

User config: ~/.engrm/settings.json

Created by engrm init. Contains API credentials, sync settings, search preferences, secret scrubbing patterns, and Sentinel configuration.

Project config: .engrm.json (optional)

Place in your project root to override project identity for non-git projects:

{
  "project_id": "internal/design-system",
  "name": "Design System"
}

Claude Code integration

Engrm auto-registers in:

• ~/.claude.json — MCP server (engrm)
• ~/.claude/settings.json — 6 lifecycle hooks

Tech Stack

• Runtime: TypeScript, runs on Bun (dev) or Node.js 18+ (npm)
• Local storage: SQLite via better-sqlite3, FTS5 full-text search, sqlite-vec for embeddings
• Embeddings: all-MiniLM-L6-v2 via @xenova/transformers (384 dims, ~23MB)
• Remote backend: Candengo Vector (BGE-M3, Qdrant, hybrid dense+sparse search)
• MCP: @modelcontextprotocol/sdk (stdio transport)
• AI extraction: @anthropic-ai/claude-agent-sdk (optional, for richer observations)

License

FSL-1.1-ALv2 (Functional Source License) — part of the Fair Source movement.

• Free to use, modify, and self-host
• You cannot offer this as a competing hosted service
• Each version converts to Apache 2.0 after 2 years
• Sentinel is a separate proprietary product

See LICENSE for full terms.

Built by the Engrm team, powered by Candengo Vector.