RAN Intelligence · Agent access

Take the evidence
into your own workspace.

Connect an approved AI client to private project evidence, source-backed Roof IDs, Bid Truth assessments, Decision Receipts, the published Tie-In record and current market signals. Private customer-system context stays off unless both an owner-approved field grant and an opted-in token permit it.

Same permissions

An agent can only retrieve private records available to your company workspace.

Evidence boundaries

Official source proof, transcript context and labeled inference remain distinct in every market result.

Revocable access

Tokens are hashed at rest, expire automatically, and can be revoked here immediately.

Secure sign-in

Manage agent access.

Use the same work email as Project Workspace. We will send a password-free sign-in link.

Connection

One endpoint. Thirty-seven evidence tools.

Paste the token above into an MCP client that supports a remote URL and custom authorization header.

{
  "mcpServers": {
    "ran-project-intelligence": {
      "url": "https://www.roofingauthority.network/api/mcp",
      "headers": {
        "Authorization": "Bearer ran_mcp_YOUR_TOKEN"
      }
    }
  }
}

Project evidence: projects, documents, revisions, findings and review lineage.
Industry and market evidence: transcripts, signals, official-source changes and publisher outputs.
Operator evidence: Decision Receipts and human-approved learning history.
Roof identity: list Roof IDs, inspect one roof, retrieve its source-backed facts, and enumerate unresolved unknowns.
Bid Truth: inspect pursuit economics, evidence gaps, risk-adjusted assessments and permanent human bid receipts.
Private context: approved normalized projections only; raw payloads and credentials remain outside MCP.

Read the integration guide →