---
title: "Designer MCP"
description: "Let an authorized external agent build and review one chart through SeatLayer's deterministic remote MCP workflow."
---

The remote Designer MCP lets Claude, ChatGPT, another MCP host, or your own agent
operate a chart-scoped authoring workflow. SeatLayer supplies deterministic
tools, validation, rendering, and review gates; the connected host supplies the
model.

Choose a setup guide for [Codex](/agents/connect-codex),
[Claude](/agents/connect-claude), or [ChatGPT](/agents/connect-chatgpt), and
review the [current capability matrix](/agents/capabilities) before enabling tools.

<Aside type="caution" title="Environment-dependent access">
  Availability, connector URL, and publish permission are shown on the Agent
  connections page in your SeatLayer dashboard. Do not copy a development URL
  into production or present this workflow as a one-click image scanner.
</Aside>

## Security boundary

OAuth authorizes one chart-scoped session. The agent cannot use that token to
enumerate unrelated charts, workspaces, events, members, buyers, billing, or API
keys.

Compatible hosts need:

- remote Streamable HTTP MCP;
- protected-resource discovery;
- OAuth 2.1 authorization code flow with PKCE S256;
- the `resource` parameter and rotating refresh tokens; and
- MCP image content for reference and preview inspection.

Never give the model a SeatLayer secret API key.

## End-to-end workflow

<Steps>
  <Step title="Negotiate capabilities">
    Call `get_capabilities` and follow the features reported by this deployment
    instead of assuming a tool or schema from memory.
  </Step>
  <Step title="Authorize one chart">
    Select an existing chart or create an empty scoped draft during OAuth.
  </Step>
  <Step title="Upload and classify private evidence">
    Keep immutable asset ids, inspect overview/detail tiles, classify every
    source, and configure one trace-ready geometry source plus role-limited
    supplemental evidence.
  </Step>
  <Step title="Confirm semantics">
    Record region meaning, labels, counts, confidence, and open questions.
    Pixel inference remains evidence until an authoritative person/source
    confirms sellable facts.
  </Step>
  <Step title="Compile deterministically">
    Select semantic intents and opaque region ids. SeatLayer code—not model
    coordinates—creates sections, paths, rows, aisles, and inventory.
  </Step>
  <Step title="Render and refine">
    Validate every floor, inspect buyer previews and reference diffs, make a
    bounded change, and repeat.
  </Step>
  <Step title="Review the exact revision">
    Prepare a packet bound to the current immutable chart hash. Inspect desktop
    and mobile buyer scenes plus reference evidence.
  </Step>
  <Step title="Publish only an approved unchanged revision">
    Record the human decision. A `canPublish` session may publish only the exact
    approved revision; any change invalidates that approval.
  </Step>
</Steps>

## Why it is not an image converter

A single vision response cannot guarantee:

- authoritative seat count or labels;
- correct accessibility and pricing;
- containment-safe geometry;
- readable buyer behavior at several zoom levels;
- exact revision review; or
- safe publication.

The MCP therefore separates evidence, semantic confirmation, deterministic
geometry, machine validation, visual review, and publication.

## Start with this prompt

```text title="designer-agent-prompt.txt"
Use the SeatLayer Designer connector to build this seating chart with me.

Start with get_capabilities and follow the staged reference workflow.
Treat every uploaded image as evidence, not authoritative seat inventory.
Inspect the overview and all required detail tiles before interpreting a dense plan.
Record unresolved labels, counts, accessibility, and pricing as questions.
Use semantic tools and SeatLayer's deterministic compilers; do not invent coordinates.
After each bounded change, validate and render every floor.
Show me the buyer preview and reference alignment evidence.
Prepare the exact revision for review, but do not approve or publish without
my explicit decision.
```

## Agent implementation rules

- Use the newest `expectedUpdatedAt` on mutations to avoid overwriting concurrent
  edits.
- Split large semantic region sets into bounded, disjoint calls and read back
  the merged canonical state.
- Do not let a supplemental source replace the configured geometry source.
- Do not create inventory from unconfirmed image counts or labels.
- Prefer semantic intents; public MCP tools intentionally exclude arbitrary
  geometry nudges and several Designer-only controls.
- Treat machine validation as necessary but insufficient for visual approval.
- Preserve published booking identities and safe-mode constraints.

## Human review gate

Before approval, verify:

1. every floor and buyer scene is present;
2. source overlays/diffs match the intended geometry;
3. labels, capacity, accessibility, categories, and prices are authoritative;
4. no private reference appears in buyer output;
5. mobile and desktop interactions are readable;
6. the revision hash still matches the prepared packet; and
7. the reviewer understands the effect on new and existing events.

## Definition of done

- [ ] OAuth scope is the intended chart and workspace.
- [ ] Source classification and required tiles are complete.
- [ ] All inventory facts are confirmed.
- [ ] Deterministic validation passes on every floor.
- [ ] Buyer 2D/3D and mobile/desktop evidence is reviewed.
- [ ] Approval is attached to the exact unchanged revision.
- [ ] Publication is an explicit human-authorized action.
- [ ] The handoff records assumptions, open questions, and the published version.

Continue with [reference plan import](/designer/reference-plan-import),
[sections and seat plans](/designer/sections-and-seat-plans), or
[publishing and versioning](/designer/publishing-and-versioning).