Context7
by upstash
Feeds AI the latest docs for libraries and frameworks before it answers, so code it writes does not use outdated, removed functions.
Computer or cloud★ 64,224MITUpdated today
Install
Monet
Overview
Description from the maker (original English)
Local-first, state-centric memory for coding agents. MCP server + CLI.
PopNx MCP’s three pillars
These three are never merged into one score: opinions, measurements and popularity are different things.
- User voiceiNo reviews yetReviews open together with accounts in a later step
- Our basic checksi1 of 1 passedChecked from public information; code not scanned yet
- UsageiNot measured yet★ –
What you can do with it
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
- Connection: Runs on your computer
- Transport: stdio
- Category: Search & AI memory
Install
Pick the AI app you use, then copy the settings.
Before you install
- Basic checks: 1 of 1 passed
- Runs on your computer with your user’s permissions (files, network, apps).
- You fill in yourself:
MONET_STORAGE_DIR
You need
- npmNode.js installed (for the npx command)
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Open Claude Desktop → Settings → Developer → Edit Config, and paste this into claude_desktop_config.json
{
"mcpServers": {
"monet": {
"command": "npx",
"args": [
"-y",
"@team-monet/monet",
"start"
],
"env": {
"MONET_STORAGE_DIR": "your-value"
}
}
}
}Run this in your terminal
claude mcp add monet --env MONET_STORAGE_DIR=your-value -- npx -y @team-monet/monet startFill in the empty values before running it.
Or copy the settings yourself
Paste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json
{
"mcpServers": {
"monet": {
"command": "npx",
"args": [
"-y",
"@team-monet/monet",
"start"
],
"env": {
"MONET_STORAGE_DIR": "your-value"
}
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Or copy the settings yourself
Paste into .vscode/mcp.json in your project
{
"servers": {
"monet": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@team-monet/monet",
"start"
],
"env": {
"MONET_STORAGE_DIR": "your-value"
}
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This one runs on your computer, and ChatGPT can only connect to servers on the internet. Pick another app.
Your choice stays on this device in a cookie; no account needed.
Values you fill in
MONET_STORAGE_DIRenvDirectory where Monet stores local-first memory. Defaults to ~/.monet (organized per project by circle).
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
What it can do
Runs on your computer; the tool list appears in your AI app after you install it.
Basic checksi
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 7, 2026
- ✓ Pass
Maker matches the repo
The registry name was verified with the GitHub account team-monet, which owns the repo
- … Not checked yet
Still maintained
We have not read this repo from GitHub yet
- … Not checked yet
Has a licence
We have not read this repo from GitHub yet
- – Not applicable
Endpoint responds
This one runs on your computer; there is no internet endpoint
- – Not applicable
Sign-in with OAuth
Runs on your computer and uses local settings instead
- … Not checked yet
No hidden instructions in tool descriptions
Runs on your computer; we can only read tool lists from internet endpoints
Compared with others in Search & AI memoryi
- More stars than 0% of the other 7259 servers in this category
- Published to the registry more recently than 96% of the other 7259 servers in this category
- Basic checks equal to or better than 100% of the other 7259 servers in this category
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
Common questions
Is Monet safe?
It passed 1 of 1 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
What keys or codes does it need?
The registry does not list a required key, though some servers ask you to sign in on first use.
Does it run on my computer or in the cloud?
It runs on your computer. Your AI app starts it when needed, and data does not pass through the maker’s servers (unless it calls an online service itself).
Is it still maintained?
The GitHub repo last changed –, and the latest version (1.12.0) reached the registry yesterday.
Which apps can use it?
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need Node.js on your computer.
What can I use instead?
In the same category: Context7, Codebase Memory, Exa
Alternatives
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
Codebase Memory
by DeusData
Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
Runs on your computer★ 48,148MITUpdated 20 days ago
Install
Exa
by exa-labs
Connect AI agents to Exa for web search, content fetching, and multi-step research.
Runs in the cloud★ 5,213MITUpdated last month
Install
README from the maker
Open the README
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
monet
Memory should change what your agent does.
🌐 [team-monet.com](https://team-monet.com/) · listed on the [MCP Registry](https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.team-monet/monet)
Monet is local-first memory for coding agents, built for the outcome you meant. It keeps what you and your team actually want — principles, rules, corrections — and delivers it while the agent works: principles always on, rules read at the moment they bind, corrections recorded so they never need making twice. It runs 100% locally (SQLite + on-device embeddings) and speaks MCP.
Get started — let your agent set it up
The easiest way to use Monet is to let your agent install and wire it for you. Paste one line into Claude Code (more agents coming soon):
Set up Monet for all my projects: read https://raw.githubusercontent.com/team-monet/monet/main/harness/bootstrap/install.md and follow it, checking with me at each step.
Your agent installs Monet, wires up the MCP server, and pulls in a memory-aware agent team — no manual config, no environment variables. That's the with-monet harness, which lives in harness/ in the Monet repo, and it's the recommended path.
⭐ If Monet helps, [star Monet on GitHub](https://github.com/team-monet/monet) — it's the best way to support the project.
Zero config
Monet keeps one local store at ~/.monet and automatically organizes each project into its own *circle* — nothing to set up. Everything stays on your machine.
Run as many agent sessions as you like: they share one copy of the embedding model, so another session costs tens of megabytes rather than gigabytes. The first session that needs the model starts one background process (model-host) to hold it. It is reachable only by your own user, never over the network, and it exits by itself after a few idle minutes; the next session that needs it starts it again. It works through one request at a time, so a very long write makes other sessions wait. If it ever stops answering, the waiting call fails after ten minutes and says which process to end.
See your memory — monet dashboard
Run one command to open a local web view of your Monet memory as an explorable, force-directed graph:
monet dashboardIt opens in your browser automatically — if it doesn't (some Windows/headless setups), the terminal prints the URL (http://localhost:7373) and the store path; just visit that URL. The Graph tab is the main view: nodes are concepts (colored by kind, sized by importance), edges are the relationships between them, and a detail panel lets you read any concept's full body and evidence. The Concepts, Entities, Timeline, and Health tabs are sortable tables — Health surfaces contradictions, duplicates, and low-confidence concepts. If the graph looks empty, check the Store: line in the terminal: an empty graph usually means it's pointed at the wrong store (use -d) or the store is just new.
100% local. Fully offline. Strictly read-only. The server copies your store into a temporary snapshot on each request — the live database is never written to or locked. No network connections are made; run it with wifi off and the graph still renders.
monet dashboard # opens at http://localhost:7373monet dashboard -p 8080 # -p, --port custom port monet dashboard -d ./path # -d, --dir point at a specific store
Diagnose and repair an embedder store — monet doctor / monet repair
If Monet reports an embedder pin mismatch, an interrupted embedding migration, malformed vectors, or an unavailable local model, start with the read-only diagnostic command:
monet doctormonet doctor --check-provider # also load and probe the exact pinned provider monet doctor --dir ~/.monet --json # machine-readable result; diagnostics stay on stderr
doctor prints the absolute database path, SQLite integrity and schema, the durable model pin, all four live embedding populations (including dimensions and malformed-row samples), any migration sentinel, and copy-paste next commands. It does not construct MonetCore or alter the database. A healthy diagnosis exits 0; a completed diagnosis that needs recovery or provider action exits 2; inspection failures exit 1. An unavailable provider is reported separately from store safety—for example, an unsupported hashing tokenizer or missing ONNX cache does not by itself mean the database is corrupt. For an ONNX or custom model ID whose shape cannot be proven from the pin alone, --check-provider can reconcile an otherwise unknown assessment only when the exact provider loads and every clean live population has the same matching vector width.
When the server will not start at all
A host that cannot start the server reports only that the connection closed — the transport does not exist until every fallible startup step has already succeeded, so there is no protocol channel left to explain a failure. The cause is written beside the store instead, at monet.db.startup-failure.json: the step it died in, the error and its code, the time, and the pid. The record is a sidecar of one database, named after it, so a directory holding more than one store never mixes their diagnoses. doctor prints that record on stderr before anything else, including when the store itself is too damaged to inspect.
The record holds only the most recent failure, and a later successful start does not clear it — hosts retry, so fail → fail → succeed is ordinary, and clearing on success would erase the evidence of the attempts that failed. Read the timestamp before acting on it.
Repairs are previews unless both confirmation flags are present. Choose exactly one mode:
# Preview a complete rewrite to a built-in provider alias.monet repair --target hashing monet repair --target onnx
Exact persisted model IDs are also accepted (quote them in scripts).
monet repair --target 'hashing:dim=256:tok=2' --json
Apply only after reviewing the preview; this command never prompts.
monet repair --target hashing --apply --yes
An interrupted migration must use the sentinel's exact target.
monet repair --resume monet repair --resume --apply --yes
Abandon is available only when core proves no vectors were rewritten and can restore the prior pin.
monet repair --abandon monet repair --abandon --apply --yes
Preview classifies the durable state before offering an apply command. An empty unpinned store or a clean, provider-compatible same-target store reports action: none; no backup or rewrite is needed, and --apply --yes refuses. Abandon preview also refuses unless a sentinel exists and core diagnostics classify it as safe. A valid target or resume preview loads and probes the exact provider but does not open the database for mutation. Apply preflights the provider again, takes exclusive SQLite ownership, and creates a verified backup before core can migrate schema or embeddings. Backups are automatic and cannot be disabled or redirected:
<storage-directory>/backups/monet-before-repair-<UTC>-<uuid>.dbThe backup path is printed immediately and retained if later work fails. Stop other Monet processes if exclusive ownership cannot be acquired, then rerun the same command. Do not delete an active migration sentinel or edit the pin manually: use --resume, or use --abandon only when its preview says core classifies abandonment as safe. Successful previews and applies exit 0; invalid flags, provider failures, locks, unsafe abandon attempts, backup failures, and migration failures exit 1.
License
Monet is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0-only). Full terms are in the LICENSE file included in this package. The with-monet harness is licensed separately under the Apache License 2.0.