marketer.law for Agents

What your agent can do

These are the actual capabilities Story publishes to agents. Each one is a separate permission on the key you issue, so you can hand an agent the read-only ones and nothing else.

• ✓Start a legal process. Kick off a client retainer, campaign SOW, creator agreement, or vendor form from the same lawyer-maintained process a Story attorney would run — not an ad hoc draft against a blank page.
• ✓Create a document. Produce the draft inside your legal file, where it is versioned and findable, instead of as an attachment in someone's inbox.
• ✓Check where a matter stands. Read the status of a transaction in flight, so your agent stops asking you whether the client signed.
• ✓List what is outstanding. Pull your pending legal tasks and surface them wherever your team already works.
• ✓Read pricing, usage, and your subscription — and, if you grant that permission, purchase one.

Mapped to the month you actually have

What your agent cannot do

This list is the reason to trust the rest of it. These are limits enforced by the platform, not guidelines an agent is asked to follow:

1. 1)Execute a signature. Signing and irreversible legal execution sit outside the agent surface entirely. Your agent can get a document to signature-ready. A person signs it.
2. 2)Close out your legal tasks. An agent can read what is outstanding; it cannot write back that the work is done.
3. 3)Reach outside your organization. Access is scoped to your organization at the middleware layer, so an agent cannot read another company's data regardless of how it is prompted.
4. 4)Start a process you did not allow. Every key carries a whitelist of the specific processes it may initiate. Anything outside it is refused, not improvised.

How it actually connects

Four steps, and if anyone on your team has configured an MCP server before, they already know three of them:

1. 1)Switch it on — agent execution is off until you activate it in your organization settings. Nothing can happen before you do.
2. 2)Authenticate — with a Story-issued API key, scoped to your organization and to the permissions you pick.
3. 3)Connect — point your agent at the Story MCP endpoint.
4. 4)Run the work — start processes, create documents, and read status, with a person still on every signature.

{
  "mcpServers": {
    "story": {
      "url": "https://backend.story.law/mcp/story-agent",
      "type": "http"
    }
  }
}

Prefer plain text? This page is also available as markdown — built for feeding directly to your own agent.

Every action an agent takes is logged in an audit trail that distinguishes programmatic actions from human decisions — which matters later, when a client asks who approved what.