Revealed MCP server icon

Revealed

by Revealed

HIPAA CompliantSOC2 ReadyISO 27001 Ready
Sales & CRM14 tools

Read a healthcare-sales account book, buying-group people, change signals and ranked next actions from an AI agent. 14 tools, all read-only, with five read-only OAuth scopes.

Verified connector

Listed by Anthropic as a partner connector in its Connectors Directory.

Connection checked by Agentman on .

Anthropic states this reflects the level of review a connector received, not a security audit.

Connect Revealed via MCP

https://mcp.revealed.ai/mcp

Works in any MCP-compatible client. In Agentman Studio it is one click — no config file to edit.

Revealed Tools & Capabilities (14)

Tool names from Anthropic's directory listing. This server requires sign-in, so we could not read tool descriptions or parameter schemas.

Limits

  • Revealed names no tool anywhere public. The 14 names come from Anthropic's directory. The vendor documents five read-only capability categories instead, which constrains the surface but matches it only by semantic label — so a renamed tool would not show up in any diff.
  • We could not read tool schemas or safety annotations. The endpoint returned 401 to an anonymous request, so no parameter-level detail, and no server-declared readOnlyHint or destructiveHint, is published here.
  • The scopes offer no internal boundary. All five are read scopes and all five are requested together. You cannot grant account reading without also granting access to named individuals' profiles.
  • You only see accounts you pay to monitor. The free tier covers two accounts. The connector reads the curated store, so an account nobody provisioned is simply absent — a question about it returns nothing rather than triggering research.
  • Adding accounts is expensive and immediate, though not via MCP. A tier increase is charged prorated on the day it happens, Overage is billed unprorated in arrears, and tier decreases wait until the next period with no refund.
  • Accuracy is disclaimed. Revealed's terms provide the Service "as is", disclaim all warranties including that AI output will be accurate or complete, and place responsibility for reviewing output on the customer — notwithstanding the 95% accuracy figure in its marketing.
  • The skills Revealed ships are outside this audit. Its ten published Claude skills, and the contracted Automations available on plans of 50 or more monitored accounts, run inside your own AI environment and borrow whatever your other connectors can do. This page characterises the 14 tools, not the assembled workflow.
  • A signal's freshness is not knowable from the tool name. Revealed says it continuously refreshes and alerts on meaningful change, but no listed tool reports when a given record was last verified.

Frequently asked questions

No. All 14 tools read, all five OAuth scopes end in ':read', and Revealed's privacy policy states the integration is read-only and that it does not receive or store data from your CRM, email or calendar. The connector moves data one way, out of Revealed and into your agent.

No individual call spends anything. Revealed bills per 'Accounts Monitored' per billing period, not per query, and no listed tool adds an account. The 'plan_usage' tool lets an agent read your plan standing before you act, so the surface can observe its own metering.

Named healthcare executives who are not your customers. Revealed profiles decision-makers such as chief medical officers and builds records from LinkedIn posts, conference appearances, awards and publications. Its privacy policy calls this a contributor database and offers those individuals an email opt-out.

No tool names, but yes to a capability list. Revealed's help centre documents five read-only categories the assistant can access and states it cannot write, edit or delete anything. That negative enumeration maps one-to-one onto the five OAuth scopes, so the surface is constrained from two directions.

Yes. The server publishes a dynamic client registration endpoint at /register and advertises the authorization code grant with PKCE, which is what Claude Code requires. Sign-in is a magic link emailed to you, or a handoff from an existing app.revealed session.

A signal is a dated change on an account or person, such as a new chief financial officer, an RFP, a partnership or a conference panel. A persona is a buying-group role like clinical champion or IT approver. Tools read both.

No, and Revealed says so itself. It markets 95% research accuracy and says humans validate results, but its terms provide the service as is, disclaim all warranties that AI output will be accurate or complete, and state that customers must review output before relying on it. Treat retrieved signals as leads to verify.

Yes, by composition. Revealed publishes ten Claude skills that pair its read-only data with your CRM, Gmail, Slack and outreach connectors, and its outreach skill tells RevOps teams to schedule runs and post drafts into Apollo, Outreach or Salesloft. The sending capability is those connectors, not Revealed.

Sources

Use in Agentman

Connect once and your agents call these tools on their own — on a schedule, in a workflow, with nobody at the keyboard.

Open in Agentman Studio

Resources

Server Info

Category
Sales & CRM
Developer
Revealed
Tools
14
Domain
mcp.revealed.ai

Using Claude Desktop or another MCP client? Setup docs — the connection URL above works anywhere.

Similar MCP Servers for Sales & CRM

HubSpot logo

HubSpot

by HubSpot

Coming Soon
Sales & CRM5 tools
View detailsConnect
ZoomInfo logo

ZoomInfo

by ZoomInfo

Coming Soon
Sales & CRM10 tools
View detailsConnect
Explorium logo

Explorium

by Explorium.ai

Sales & CRM10 tools
View detailsConnect

Related Resources

Browse All MCP ServersClaude & ChatGPT Setup GuideBrowse AI Skills