The kōdlo MCP connector:
your GTM engine inside
any AI agent
kōdlo is a GTM engine and workspace with its own remote MCP server. Connect it to any AI agent that supports MCP, and your campaigns, contacts, research and Knowledge Hub are there in the chat: the agent reads them, updates them and starts research runs as you, with your role's permissions. It works on every plan, the Free plan included.
Free plan · no credit card required · no time limit
Who in EU fintech hasn't heard from us in two weeks, and what was the last touch?
Three people have had no touch for two weeks or more: Ruben de Vries, a LinkedIn message 19 days ago1; Aoife Brennan, a call 17 days ago2; Tiago Moreira, an email 15 days ago3. Want a follow-up ticket for each?
- 1 kodlo_get_timeline Timeline
- 2 kodlo_get_timeline Timeline
- 3 kodlo_get_timeline Timeline
What is an MCP connector?
MCP, the Model Context Protocol, is the open standard AI agents use to reach other software. An MCP connector gives an AI agent another product's data and actions during a chat. kōdlo's connector is a remote MCP server at api.kodlo.ai/mcp: add it once, sign in, and any chat can read and update your sales workspace.
An AI agent can work with your pipeline in two ways.
-
Copy and paste
You export a list, paste it into the chat and copy the answer back by hand. The agent never sees what changed since, and nothing it writes reaches your workspace.
-
kōdlo takes this road
Connected
The agent works in the workspace itself through 47 tools: it reads campaigns and contacts, writes changes back, runs research and files the results where the team can find them.
Built for teams that already work with AI agents
Founders, sales teams and agencies who research, write and plan with an AI agent, and want their pipeline in the same conversation instead of in another tab.
Five things that break when your AI agent cannot see your pipeline
-
01 Today
The agent doesn't know your pipeline
You paste a CSV into the chat, and by Friday it is out of date.
01 With kōdloThe agent reads campaigns, contacts and timelines as they are now, and writes changes back.
-
02 Today
Research stays in one person's chat
A good account summary lives in one teammate's chat history.
02 With kōdloA deep research run files a cited report in the workspace, where the whole team can read it.
-
03 Today
Your documents are out of reach
Price lists and case studies sit in a folder the agent can't open.
03 With kōdloThe agent searches your Knowledge Hub and adds new documents to your own projects.
-
04 Today
New leads mean another tool
Finding prospects means switching to a database and importing a file.
04 With kōdloThe agent runs a lead discovery against your ICP, reviews the candidates and imports the ones you accept.
-
05 Today
Follow-ups slip
Who went quiet is a question nobody has time to answer.
05 With kōdloAsk in plain words; the agent checks the timelines and opens a ticket for each follow-up.
How to connect kōdlo to your AI agent
-
Turn on
Turn on external AI apps.
The workspace owner does this once: in kōdlo, open Settings → AI apps (MCP) and switch External AI apps on.
-
Add
Add the connector.
In your AI agent's connector or MCP settings, add a custom connector with the server URL https://api.kodlo.ai/mcp.
-
Approve
Sign in, or use a key.
Agents that support sign-in open kōdlo: choose full access or read-only and click Allow. Others, such as coding agents, use an API key from Settings → AI apps (MCP).
-
Ask
Ask in plain words.
The agent picks the tools. Lead discovery and deep research run in the background, and each run also appears as a conversation in your kōdlo archive.
-
Review
Keep the final say.
Deleting contacts or campaigns, Do Not Contact, billing and members stay in the kōdlo web app. Every change made through the connector is kept in an audit trail for 90 days.
You decide how much the agent can do
The connector acts as you: it sees and changes only what you could in the kōdlo web app, in your workspace, with your role. You choose full access or read-only when you approve, and you can disconnect at any time. Agents kōdlo can't identify are marked as unverified, with the address the approval goes to.
- Workspace
- Northbeam
- Your role
- Member
- Access
- Full access · or read-only
- Stays in the web app
- Deleting records, Do Not Contact, billing
- Update contacts and campaigns
- Run lead discovery and deep research
- Add Knowledge Hub documents
- Create tickets and log notes
-
DeleteStays in the web app
What your AI agent can do with kōdlo
An AI agent on your desktop that can use your browser already writes emails and messages. Connected to kōdlo, it also knows who it is writing to and remembers what it did: it runs deep research on the profile, writes the email in your mailbox or the WhatsApp message in WhatsApp Web, logs the touch on the contact's timeline, and keeps all of it in one shared memory layer that your team and their agents work from. You still decide what gets sent.
| Copy and paste into a chat | An agent with your browser | The same agent with kōdlo | |
|---|---|---|---|
| 01 Writes an email or a WhatsApp message | A draft you copy into your mailbox yourself | In your browser, from what you tell it | In your browser, from the contact's record, the latest research and your Knowledge Hub |
| 02 After the message is written | You log it by hand, if you remember | Nothing is recorded anywhere | The touch is logged on the contact's timeline in kōdlo |
| 03 Researches a person or a company | Works only with what you paste in | Browses the web; the answer stays in one chat | A deep research report with sources, filed where the whole team can read it |
| 04 Remembers what happened | Only what is in this chat | Only your own chat history | Calls, meetings, emails and notes on one timeline per contact |
| 05 Works with your team | Each person pastes their own context | Each agent starts from zero | One shared memory layer for the team and their agents, each with their own permissions |
| 06 Follows up | When you remember to ask | Only while the chat is open | Tickets with an owner and a deadline, and a quick answer to who went quiet |
Questions about the connector
Any AI agent that supports remote MCP servers, with sign-in or an API key. The docs have step-by-step guides for the most popular ones.
Setup guideYes, through kōdlo's MCP connector. The agent can read and update contacts and campaigns, run lead discovery and deep research, search your Knowledge Hub and manage tickets: 47 tools in all.
Tools referenceYes. It works on every plan, the Free plan included. Runs that use AI, such as lead discovery and deep research, draw on your plan's AI credits.
See pricingOnly for agents that can't sign in, such as most coding agents. Create one in kōdlo under Settings → AI apps (MCP) and send it as Authorization: Bearer <your key>. It acts as you, and you can make it read-only or let it expire.
kōdlo itself never sends anything. An AI agent that can use your browser can write the email in your mailbox or the WhatsApp message in WhatsApp Web, and log it on the contact's timeline through the connector. What gets sent, and by whom, stays your decision.
Delete contacts or campaigns, set or lift Do Not Contact, edit the workspace profile, or change settings, billing or members. Those stay in the kōdlo web app, for every role.
Only what you could see in the kōdlo web app, in your workspace, with your role's permissions. A tool's result becomes part of your conversation and is processed by the agent's provider under your account's terms.
Privacy & dataGive your AI agent your pipeline
Sign up, connect kōdlo to your AI agent in a few minutes, and ask about your accounts in plain words.