Revenue Scout MCP

Find and vet public sales conversations, see why a lead fits, and prepare replies for manual review in Claude or Codex.

Server MCP Terhosting

npx add-mcp 'https://revenuescout.ai/api/mcp'

Terpasang ke Claude Code, Codex, Cursor, dan lainnya

Dokumentasi

From your conversation.To the conversation that matters.

Follow one fictional request for a bookkeeping business. Reading the saved lead starts no new search.

01 / YOU ASK

Bring me the context.

“For Example Bookkeeping, show the latest vetted leads about Shopify payouts. Explain why they fit.”

The agent resolves the permitted brand and reads its saved leads and available source evidence.

02 / SCOUT FINDS THE SAVED WORK

Reddit Example source

“We spend days matching Shopify payouts to orders. Can anyone recommend a bookkeeper?”

Why it fits The post describes the reconciliation problem in your bookkeeping offer.

Original-post context and Scout’s assessment remain attached to the lead.

03 / YOU CHOOSE THE NEXT STEP

Prepare a reply.

“Draft a useful reply to this lead using our bookkeeping offer.”

Shopify payouts can combine several orders, fees and refunds. Start by matching each payout to the transactions included in it…

Illustrative draft, ready for your review

You check the facts, edit the draft and post from your own social account.

The next request carries a learning signal. Initial reply interest and recorded results inform related search-phrase exploration. Your authored phrases stay under your control.

See replies and learning

Give Scout a job.Keep the decisions that matter.

Use these examples with your own brand and permitted records. Your key, role and current account terms determine the work available.

01

Find new leads once

“Prepare a Finder search for teams asking for help with Shopify payouts.”

Scout saves the request, returns current campaign and offer choices, then prepares the exact phrase, platform and credit review. You explicitly authorize Start. Ask for that request’s status and results.

Review and Start are separate. An uncertain response is recovered using the same saved request.

See Finder

02

Get more profile context

“Preview the available profile lookup for this lead and show the cost.”

Scout returns source-bound targets and a current preview. After your authorization it prepares the lookup and makes its saved status and permitted evidence readable.

A lookup does not establish a verified contact or a qualified lead. Reuse the same request after an interrupted acknowledgement.

See Scout’s full job

03

Keep the pipeline current

“Find Anna in this brand’s team and assign this lead to her.”

The agent discovers current permitted team choices, asks when labels are ambiguous and updates the chosen record against its current version. It reports the actual saved result.

Existing active automations can react to a sales change. Eligible undo uses the saved event and current record.

See the pipeline

04

Prepare a connected handoff

“Find our Sales channel connection and draft a rule for newly qualified leads.”

Scout lists permitted connection names, adapters and retained states. Your agent saves a draft and can preview an existing event. New connections and activation use the dashboard.

A connected state is the saved state, not a fresh provider health check. Editing a rule clears its prior activation approval.

See automations

05

Work with one reviewed Gmail email

“Prepare one email for this lead using my permitted mailbox. Show me everything before sending.”

Scout resolves the available mailbox and exact source-backed recipient. You review the message, timing and cost ceiling. Approval and queueing are separate actions for that individual email.

No inbox reading or automatic follow-up. An uncertain send requires checking the existing send rather than sending another copy.

See the Gmail connection

06

Check what Scout has done

“Show the saved discovery activity and the API capacity this key can read.”

Read permitted run, monitoring and usage state with its dates and provenance. Ask for the same saved task when you want to check progress.

A retained status is not a new platform scan. Recurring schedules are managed in the dashboard.

See recurring discovery

Get Scout workingwith your agent.

Prepare the business, choose access and try a saved read. You can finish signup, verification and consent yourself.

01 / PREPARE YOUR BUSINESS

Choose the access.

Start free or sign in. Review your business and campaign setup, then open API & MCP in Settings.

Choose a task, brands, permissions and a lifetime from the current account policy. Review the exact key before creating it. Save the one-time secret privately.

Open API & MCP

Your agent receives only the brands and permissions you review. Additions need a new key review.

02 / CONFIGURE YOUR CLIENT

Connect your agent

Make your key available to the client as REVENUE_SCOUT_API_KEY in your private environment. Run this command, then open /mcp to check the connection.

Official setup guide

The configuration contains placeholders, not your key. Save the key privately and supply it only to the client you chose. Configuration copied means the text was copied; check the connection in your client.

First request: “List my permitted brands in Revenue Scout, then ask which brand’s saved leads I want to read.”

03 / ASK FOR A SAVED READ

Start with the right brand.

“List the brands I can access in Revenue Scout, then ask which brand’s saved leads I want to read.”

Choose the correct brand. Ask for a vetted lead and its available original source. A successful read confirms that request. New paid work needs its own review.

See the exact first-read request

When both tools are available, call rs_read_capabilities with {"input":{}} to list your permitted brands. Confirm the right brand. Replace RESOLVED_BRAND_ID with that brand's ID and use this input for rs_read_prospects.

{
  "input": {
    "filters": {
      "workspaceIds": [
        "RESOLVED_BRAND_ID"
      ],
      "qualification": "qualified",
      "pageSize": 1
    }
  }
}

This reads one qualified saved lead when available. A successful empty result means there are no matching saved leads.