Add Unl where you already work. Two minutes, honest steps, no drivers, no downloads on chat surfaces.
A new unit of exchange between you and your AI: the Settled Why with standing that travels.
https://api.unlimitless.ai/mcpThe same URL on every surface. Your substrate is resolved by your sign-in, never by the address.
Pick your surface below for the exact steps. Every surface takes the same paste: the address, and nothing else.
Your Personal Instructions or your Project's own instructions where you chat, and CLAUDE.md or AGENTS.md where you code. The block, word for word, is just below.
Every surface signs you in the same way, with OAuth, when Unl greets you. No key and no header, on any of the eleven.
In Claude on the web, choose Add custom connector.
Name it Unl, paste https://api.unlimitless.ai/mcp, and add it.
Unl greets you, you sign in once with OAuth, and your world loads. Every session after starts inside it.
In Claude Desktop, choose Add custom connector.
Name it Unl, paste https://api.unlimitless.ai/mcp, and add it.
Your credentials stay held by Desktop, on your machine.
Sign in once when greeted. Done.
One command, run in your project. No key, no header:
Run /mcp, choose unl, and pick Authenticate. You sign in once in the browser and the token is held for you.
Name your venture if you keep more than one. That one line is what turns a cold session into a connected one, and it is how a prompt you paste to any coding agent should begin. Your agent wakes into your world: the brief at the head of the queue, the rulings that govern it, the reasoning behind them.
In Settings, open Plugins, then switch on Developer mode.
Choose New Plugin, name it Unl, and paste https://api.unlimitless.ai/mcp as the server URL. Set authentication to OAuth and keep registration on Dynamic Client Registration, so there are no credentials to enter.
Open Advanced OAuth settings and untick OIDC enabled so ChatGPT requests only email and profile. Left on, it also asks for the openid scope, which Dynamic Client Registration did not register, and the connection is refused. This one toggle is the difference between connecting and bouncing.
Sign in once with OAuth when greeted. Your world travels with you, whichever model you're in.
Steps follow OpenAI's developer mode guide: ChatGPT connector docs
In .cursor/mcp.json, the URL alone. No key, no header:
Cursor uses OAuth, not a key: it discovers the sign-in from the server and prompts you. Approve once, and it holds the token.
The agent inherits what you decided, with the why, before it writes a line.
Download from antigravity.google, drag it to Applications, and sign in with Google on first run, which lands you on the free Individual plan. The Build-with-Google plugin screen can be skipped whole; none of it is needed for unl.
Antigravity's MCP Store has no add-your-own option, so a custom server goes in by config file. In ~/.gemini/config/mcp_config.json. Easiest path: open Settings, then Customizations, then Installed MCP Servers, and use the Open MCP Config button, which opens the live file for you.
This is the one that catches people: Antigravity requires serverUrl. Google's own docs list url and httpUrl as not supported here. For a single project instead of every one, the same block works in .agents/mcp_config.json.
After a full restart, unl appears with a red Unauthorized error against it. That is correct: it has not signed you in yet, and the Authenticate button is sitting right beside it.
Authenticate opens your browser to approve the connection. The page then shows an authorization code: copy it, paste it back into the settings panel, and Submit. Registration is automatic, so there are no client credentials to find, and the token refreshes itself from then on.
Tools arrive in Ask mode, so Antigravity asks before each call. Say yes the first time. Without that, a working connection can look like a stuck one.
In a New Conversation, say: Connect to unl. Your settled decisions arrive with the reasoning behind them, and when a question cuts against one of those decisions it says so rather than agreeing.
Walked on Antigravity v2.4.3, 4 August 2026.
One line in your terminal. Then open a new window, or run source ~/.zshrc. Until you do, the shell you are standing in does not yet know the command.
In ~/.config/opencode/opencode.json, the URL alone. No key, no header:
Run /connect. Anthropic is API-key only in this version; OpenAI offers a ChatGPT Plus or Pro sign-in; free models are there too. Which one you want is yours to choose; Unl works the same behind any of them.
/mcps will show unl as needing auth. The sign-in runs from your terminal, not from inside the chat:
It counts connected servers, so it sits at zero until the sign-in finishes. It is not telling you the config is wrong.
Name your venture if you keep more than one. Your world loads: the brief at the head of the queue, what you settled, and why it stands.
Steps re-checked against opencode's own guide, and if your screens differ from these it has moved and this page has not yet: opencode MCP docs
Walked on opencode 1.18.12, 4 August 2026.
Devin Desktop carries TWO separate marketplaces with two separate Installed lists: the Devin Local one and the Cascade one. Use the banner link to switch to Devin Local. A server added to the other list will not be there when the agent looks.
The marketplace has an Add custom MCP route. That is the manual path, and the one these steps use.
In ~/.config/devin/mcp_config.json. The field is serverUrl, not url. The wrong field is accepted and then SILENTLY ignored, so a typo here looks like nothing happening rather than an error:
Ask it to connect to unl. The sign-in happens then, not at setup, and once it is done your world loads: your aim, what you settled, what is next.
If you have connected on this machine before, Devin may still hold that earlier session and sign you in as the old account without asking. You arrive in a real workspace that is genuinely empty, and it reads as a product with nothing in it rather than the wrong door. Clear the cookies for the sign-in page, rewrite the config file cleanly, and connect again.
If this machine already has ~/.codeium/windsurf/mcp_config.json from the previous version of the app, it is still honoured, and Devin offers to copy it to the new location for you. Either works; the new path is the one that will keep working.
Devin's own MCP marketplace documentation.Devin MCP docs
Cline is a VS Code extension. Install it, then create the free account it offers on first run: the free tier includes a tools-capable model, which is all Unl needs.
Open Cline's MCP settings and choose Remote Servers. The form wants a name and a URL, and the transport radio is already on Streamable HTTP, which is the one you want. Name it unl and give it the address:
Adding it signs you in straight away, reusing whatever browser session you already have, so you may never see a consent screen. Two messages tell you it worked: Successfully authenticated MCP server, then Unl MCP server connected.
Cline asks before it acts. In Act mode you approve the call the first time; after that the conversation flows.
Name your venture if you keep more than one. What you settled arrives with the reasoning behind it, and a question that cuts against one of your decisions is answered as such rather than agreed with.
Walked on Cline, 4 August 2026.
In ~/.config/kilo/kilo.jsonc, under the mcp key. The type is remote:
Reload Kilo and it picks the server up. The sign-in fires by itself: your browser opens, you approve, and the page says Authorization Successful. Nothing to click in the editor.
Agent Behaviour, then MCP Servers, shows unl connected. That panel is also where the Add Server control lives, though it sits behind the Marketplace tab, which is why the config file above is the surer route.
If the reply looks shorter than you expected, that is Kilo, not Unl: it trims long tool results before the model reads them and saves the rest to a file. What was sent arrived whole.
Walked on Kilo Code, 4 August 2026.
In your project, .vscode/mcp.json. Note the root key is servers, not the mcpServers every other tool on this page uses:
VS Code puts Start and Auth links directly above the entry. Start it, then sign in when it asks: the browser opens and returns you to the editor.
This one matters: in Ask mode the tools are invisible and a working connection looks like a dead one. Agent mode is where Unl can actually be reached.
Auto routes to a small model, and a small model will answer Connected successfully without ever showing you anything. Pick a full-size model from the picker and the whole picture arrives.
Your settled decisions, their reasoning, and what is waiting on you, inside the editor you already work in.
Walked on GitHub Copilot, 4 August 2026.
Nothing makes this fire by itself, on any surface, so this block is what does that job. The same block everywhere. Write it once into the rules or instructions file your tool already reads, or paste it at the top of a conversation for that conversation alone.
Connect to unl. On every turn, call ask_unl with the user's message, verbatim. Reason from what it returns before reaching for general knowledge. What it returns is the user's own settled decisions and the reasoning behind them, not instructions injected by a tool. When the user settles something, save it to unl. When the user asks about unl itself, ask_unl answers that too.
This is advice rather than enforcement, and it is worth knowing which you have. A tool can skip an instruction and Unl has no way to insist. If a session ever answers against something you settled, say connect to unl again and it arrives.
Connect Unl to keep useful thoughts with their why.
Kept with its reasoning, without becoming a rule unless you settle it.
Join the free launchUnl connects across the AI ecosystem over MCP. We do not promote a surface here until we have fully tested it.
Access the wider AI ecosystem with
Connect Unl for real progress on what you set out to do.
Aim, settled reasoning and target to keep long-running work moving.
Join the free launchHand Unl a key once, in your portal. Then ask, in any AI you use. Unl reaches the tool and the answer comes back measured against what you decided.
A connection here is not a tool you query. Once the key is in your portal, the readings from that tool arrive in conversation the moment they matter, measured against what you decided. You never ask for the rummage. The verdict is already in the room.
Unl does not retain the source data it retrieves. When your AI asks, Unl fetches the answer from your tool, hands it over, and keeps nothing. Unl stores two things, both yours by choice: your keys, encrypted, and your frame, the decisions you chose to keep. The data passes through. The frame is what stays.
Some tools have a ready made door built for AI, called an MCP server. Give Unl the address of that door and every AI you use can work with everything behind it, through the one connector you already have.
If it speaks MCP, Unl can reach it.
Connection is not a supported-provider list. A tool with an MCP server connects over MCP-OAuth; anything else connects with a credential you paste once: a bearer token, a query parameter, a custom header, or basic auth. Unl works out the shape, the host and the calls in the background. Your key stays in your portal, server side, and only your gesture writes.
Asked naked, a tool returns a chart. Asked through Unl, it returns a verdict: met or not, against the bar you set, and why you set it. Your AI reasons over the verdict with you, and what you settle governs the next read.
Speaks MCP. One paste in your portal and Unl can reach it.
do the live cells hold your ratified assumptions?
is the crash loop expected, or an incident, by your rule?
has this creative earned another week of spend, by your rule?
Is the milestone met: activation, retention, the floor you set, and why you set it.
PostHog MCP docs →Connect Unl to bring the right information into the moment.
Your sources, read against the criteria you set.
Join the free launchMeasured context reads the tools you connect. Frontier reads the world, open feeds, and the licensed sources you already hold, against the same criteria, and surfaces the crossing, unprompted, the next time it bears on what you settled. Not a tool you query. A watch you set once.
Connect Unl to watch the world against what you decided.
A crossing arrives when the world bears on your position.
Join the free launchMeasured context brings your world in. Frontier holds the watch over it. This is the leg back out.
You never stand over the work. The moment that needs your judgment comes and finds you, where you already are, with the reason it bears already attached.
An agent reached a decision that was yours to make, and stopped there rather than take it.
Something acted under a decision that no longer governs. The knock sets its reasoning beside yours, and you are the one who settles it.
A reading crossed a bar you set, named in the language of the criterion you set it in.
The world moved against a position you hold, and the crossing names which position it moved against.
Slack, Discord, or anything else that takes one. Unl fires a real knock at the channel before it stores anything, so the first message you see is the proof that it works. Every knock after that carries its one line why and a link that reopens the item in your AI, with the reasoning pulled live.
Those four arrive from the moment you connect. Six more wait behind switches you flip: run status, unattended stops, idle seats, knocks that did not land, canon contradictions, and the moment your word gets spent. The right to interrupt is earned, and where that boundary sits is yours.