MCP Servers · Memory
HBarefoot/engram
🧠 Persistent memory for AI agents. SQLite for agent state. Zero cloud dependencies. Local embeddings. MCP-native integration with Claude Desktop/Code, Cursor, Windsurf & more.
Our verdict
We have not reviewed HBarefoot/engram yet
No one at Enterprise DNA has checked this listing against its sources. The description comes from public listings gathered automatically. Any fact below shows where it came from and when it was last checked.
From the listing
About HBarefoot/engram
Engram is a local memory system for AI agents. It uses SQLite to store agent state and embeddings, eliminating cloud dependencies. It integrates via MCP with tools like Claude Desktop, Cursor, and Windsurf.
Best for
Developers building local AI agents that need persistent memory without cloud services.
Use cases
- Store conversation history locally for AI agents
- Enable agents to recall context across sessions
- Run memory persistence without external APIs
Notes
4 stars on GitHub. Last updated 2026-07-01. Licensed MIT.
Pros
- No cloud dependencies, fully local
- Simple SQLite-based storage
- Works with multiple popular AI tools via MCP
Cons
- Limited to local machine, not distributed
- Embeddings are local, no cloud-scale retrieval
- Small community (4 GitHub stars)
Key facts
What we have checked
- Days since the last commit
- 133
- GitHub forks
- 1
- GitHub stars
- 9
- Days to a first reply on issues
- 138.8
- Open issues
- 2
- Releases in the last 90 days
- 0
Each value shows when we last confirmed it and where it came from. Values we have not confirmed are left out, never guessed.
How to use
How to set it up
As the listing publishes it. We have not run this install.
Install
npm install -g @hbarefoot/engramTools exposed
engram_rememberengram_recallengram_forgetengram_feedbackengram_contextengram_status
Clients the listing names (not checked here)
Claude Desktop, Claude Code, Cursor, Windsurf, Cline, ChatGPT
Install
Add HBarefoot/engram to your client
Our lab has not run this server yet. "Config checked" on a client means the recipe matches that client's documented config format; it does not mean we ran it.
Claude Code
Config checkedRun in your terminal (runs on your computer)
claude mcp add --transport stdio hbarefoot-engram -- npx -y @hbarefoot/engram@1.6.3 startClaude Desktop
Config checkedAdd to claude_desktop_config.json (runs on your computer)
{
"mcpServers": {
"hbarefoot-engram": {
"command": "npx",
"args": [
"-y",
"@hbarefoot/engram@1.6.3",
"start"
]
}
}
}Cursor
Config checkedAdd to .cursor/mcp.json (runs on your computer)
{
"mcpServers": {
"hbarefoot-engram": {
"command": "npx",
"args": [
"-y",
"@hbarefoot/engram@1.6.3",
"start"
]
}
}
}VS Code
Config checkedAdd to .vscode/mcp.json (runs on your computer)
{
"servers": {
"hbarefoot-engram": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@hbarefoot/engram@1.6.3",
"start"
]
}
}
}Codex
Config checkedAdd to ~/.codex/config.toml (runs on your computer)
[mcp_servers.hbarefoot-engram]
command = "npx"
args = ["-y", "@hbarefoot/engram@1.6.3", "start"]- ChatGPT: ChatGPT connects only to remote servers that need no API key.
Recipes come from the server's entry in the official MCP registry. Check the package name before you run it, and keep keys out of shared files.
Works with
What HBarefoot/engram connects to
5 clients, each with how we know and when we checked.
| Works with | Type | How we know | Checked |
|---|---|---|---|
| Claude Code | Client | Install recipe matches the client's documented formatRuns locally, version 1.6.3Config checked | Claude Code docs |
| Claude Desktop | Client | Install recipe matches the client's documented formatRuns locally, version 1.6.3Config checked | Claude Desktop docs |
| Cursor | Client | Install recipe matches the client's documented formatRuns locally, version 1.6.3Config checked | Cursor docs |
| VS Code | Client | Install recipe matches the client's documented formatRuns locally, version 1.6.3Config checked | VS Code docs |
| Codex | Client | Install recipe matches the client's documented formatRuns locally, version 1.6.3Config checked | Codex docs |
Stars and usage
How HBarefoot/engram is growing
9 to 9
Health and security
Is HBarefoot/engram looked after, and what do its setup files show
Health score
21 out of 100
| Part | Points | What we saw |
|---|---|---|
| Recent activity | 20.8 of 30 | Last push 133 days ago |
| Releases | 0 of 15 | 0 releases in 90 days |
| Issue replies | 0 of 20 | First reply in 138.8 days (median) |
| Maintainer | 0 of 15 | Owned by a single user |
| Official source | 0 of 20 | Community server |
Alternatives
Alternatives to HBarefoot/engram
| Record | Type | How we know | Checked |
|---|---|---|---|
| topoteretes/cogneeThis is an alternative to it | MCP server | Named in the listing | Not dated |
| vectorize-io/hindsightThis is an alternative to it | MCP server | Named in the listing | Not dated |
Compare
Compare HBarefoot/engram side by side
Sources
Where this comes from
- GitHubRegistry
A source shows what was published when we checked it. It does not certify security, performance or a working installation. Read our method.
Changelog
What changed
No changes recorded yet. Releases, price changes and deprecations appear here, with their dates, once we track them for this record.
Open-source & AI alternatives
Swap-in tools that solve the same job. Weigh the trade-offs before you commit.
Get the free Developer’s Field Guide
A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.
Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.