MARZO MCP

Give your AI
the context behind
the work.

Marzo brings current decisions, commitments, and open questions from meetings and Slack into your AI tools. Your agent can ask for more context as it answers questions, creates work, and updates your tools.

In Claude, ChatGPT, Cursor, and Codex. During meetings and afterward.

ClaudeExample

Prepare the customer reporting brief.

Context supplied through Marzo MCP

Context from Marzo

DECIDED
Monthly reports, split by region.
EXPECTED
First data export Tuesday.
STILL OPEN
Metric definitions and report owner.

From account reviews and #customer-ops

Trace sources
Account review · Meeting

“Monthly reports by region. Metric definitions and the report owner are still open.”

#customer-ops · Tim

“First data export expected Tuesday.”

MARZO + YOUR AI TOOLS

Put that understanding to work.

Put that understanding to work.

Put that understanding to work.

Get the answer you need, create a useful brief, or update an existing plan. Your AI can combine Marzo’s context with the records and actions in your other connected tools.

Get the answer you need during a customer call.

ChatGPTEXAMPLE

What billing option can we offer this customer? Use this conversation and our earlier discussions.

Used current and earlier conversations in Marzo
Context from Marzo

The customer accepts an annual term but needs an alternative to paying upfront. Finance has already approved quarterly billing on that term.

Your approved billing option
ANSWER WITH SOURCES

You can offer an annual term with quarterly billing. Finance has approved that structure. Ask whether it resolves the customer’s concern about paying the full year upfront.

Marzo provides the context. Your AI uses the other tools you connect to create documents, update records, and take action.

CONTEXT THROUGHOUT THE TASK

Let your agent ask as it investigates.

Let your agent ask as it investigates.

Let your agent ask as it investigates.

During a technical review, Codex investigates whether an old export endpoint can be retired. As it examines the code, it asks Marzo about the decisions and dependencies behind it.

Marzo connects the meeting in progress with relevant earlier conversations and later Slack replies. Codex can ask for more context as it works, then use the answers in its changes and tests.

See how the context changes the implementation.

CodexExample

Investigate retiring the old export endpoint. Use this technical review and earlier conversations in Marzo. Make the safe changes and test them.

Inspected the code: internal callers already use /v2/exports.

Codex asked Marzo1

“Why was /v1/exports kept after the internal migration?”

A customer integration still needed the old export format.

Google MeetTechnical review + earlier decision
Technical review · Alex
“Internal callers have moved to v2. Check the customer integration before removing the old endpoint.”
Earlier integration review · Tim
“Keep v1 until the customer jobs migrate. Their reconciliation still depends on the legacy export format.”
Codex asked Marzo again2

“Did that customer migration finish? Check the later replies.”

Reporting moved to v2. Reconciliation still uses v1.

Slack#customer-integrations · Clara
Later Slack reply · Clara
“The reporting job is on v2. Reconciliation still runs against /v1/exports. That part hasn’t migrated yet.”

Implementation

Kept the endpoint for the remaining integration.

Added a retirement guard and regression coverage for the customer’s reconciliation job.

  • Block retirement while a v1 consumer remains.src/exports/retirement.ts
  • Verify the endpoint remains available to reconciliation.tests/exports/retirement.test.ts
2 regression tests passing
Review the changes and tests
src/exports/retirement.ts
  // Called before retiring the v1 route.+ export function assertCanRetireV1(jobs: ExportJob[]) {+   if (jobs.some(job => job.version === "v1")) {+     throw new Error("Migrate v1 consumers first")+   }+ }
tests/exports/retirement.test.ts
✓ blocks retirement while reconciliation uses v1
  expect(() => assertCanRetireV1(customerJobs))
    .toThrow("Migrate v1 consumers first")

✓ keeps the legacy export available
  const response = await reconciliation.exportV1()
  expect(response.status).toBe(200)

2 passed
Migration follow-up draftedMove reconciliation to v2. Confirm completion before retiring the legacy endpoint.

GET CONNECTED

Connect Marzo to your AI tools.

Connect Marzo to your AI tools.

Connect Marzo to your AI tools.

Start with a question to answer, a brief to create, or a plan to update.

01

Bring the conversations into Marzo.

Connect Slack and capture the meetings relevant to your work.

02

Connect Marzo through MCP.

Connect Marzo to your AI tool. Add other MCP servers when the task needs their data or actions.

03

Ask for the work you need.

Describe the outcome you want. Your agent can ask Marzo for more context as it works and show the sources behind its answer.

A few practical questions.

A few practical questions.

A few practical questions.

01

Conversations captured in Marzo, including a meeting in progress and relevant earlier meetings and Slack threads. Your AI can use decisions, requirements, commitments, changed plans, and open questions, with sources.

02

Marzo includes source references so you can trace the context back to the conversations behind it.

03

You can ask questions and create work using Marzo’s context. Connect other tools when the task needs their records, documents, or actions, such as reading a CRM record or updating a Notion page.

04

Book a setup call to connect Marzo to your AI tool and start with a real task.

Bring your next task.

Bring your next task.

An answer during a customer call. A useful team update. A change to your code. Bring the work you want your AI to do.