tldraw SDK + MCP by tldraw

# Fix list: tldraw SDK + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/tldraw, the October 2026 research run, assessed 1 October 2026. Grade C, 61 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on tldraw SDK + MCP: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Security & auth, 40 out of 100, up to 10.5 more on the total

Why it scored 40: No API key. Licence keys are public and checked offline. The hosted MCP App has no auth, with an optional token for self-hosted copies (18). `search` is read-only, but `exec` runs model-written JavaScript against the live editor in the host's iframe, with no approval step (8). Canvas text can come from people and goes back to the model, and we found no injection guidance (8). No audit log (0). SECURITY.md takes reports at hello@tldraw.com with a 24-hour response target, but its supported-versions table still lists only 3.x while the current line is 5.5. No security.txt per the 30 September check, no bug bounty found (6).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 2. Payments & pricing, 35 out of 100, up to 8.1 more on the total

Why it scored 35: No payment protocol (0). Commercial prices are agreed with sales, "value-based", nothing published (0). Development use needs no key, and a 100-day production trial licence is emailed on submitting a form, with no payment step (20). An agent can use the SDK in development and the hosted MCP App with no signup, but production needs a person to request a licence (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 3. Reliability, 75 out of 100, up to 5 more on the total

Why it scored 75: Scored as an SDK. The `tldraw` npm package with React runtimes stated in the docs (20). A public `checks.yml` workflow and test suites across packages, pass state on main not checked (20). A stale bot marks issues after 150 days without activity and closes them 30 days later, and we didn't count open crash reports (12). RELEASES.md says tldraw doesn't follow semver and minor releases "may contain breaking changes", with warnings promised several releases ahead and breaking items flagged in each release note (8). v5 is a stable line (15). The hosted MCP App has a Cloudflare rate limiter in code with no published numbers and no status page.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 4. Agent ergonomics, 71 out of 100, up to 4.7 more on the total

Why it scored 71: Two model-facing MCP tools, with four more hidden from the model as app-only (25). `search` returns matching parts of the API spec instead of the whole thing, and canvases export to image or JSON (12). `exec` returns the thrown error text (10). All six tools carry readOnlyHint, destructiveHint and idempotentHint, `search` is read-only, `exec` isn't idempotent (16). TypeScript only, and the canvas needs a host that renders MCP Apps (8).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Transparency & trust, 51 out of 100, up to 4.3 more on the total

Made of editorial 45, provenance 57.

Why it scored 51: Source-available under the tldraw licence, explicitly not open source, with clear terms (15). The SDK runs in your app. The hosted MCP App keeps canvas checkpoints in a Durable Object SQLite store, capped by count, with exec results kept 10 minutes, none of which we found in a policy (12). RELEASES.md commits to warnings several releases before a breaking change (12). Licence pings are documented, but the two docs pages disagree and the code sends more than either says. Opting out means a commercial licence (6).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: tldraw.dev, registered 2022-04-11 (4 years) (7 of 15)
- Endpoint on the vendor's domain: tldraw-mcp-app.tldraw.workers.dev is not on tldraw.dev (0 of 15)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

## 6. Schema & documentation, 79 out of 100, up to 3.4 more on the total

Why it scored 79: Full TypeScript types, and the MCP App's `search` tool queries an extracted Editor API spec. Both model-facing MCP tools have zod inputs (22). llms.txt per the 30 September check (10). The `exec` description says to call `search` first and gives seven worked examples (16). `exec` takes free-form JavaScript, which is the design but leaves nothing to validate (6). Many examples, errors come back as JavaScript exceptions (10). Dated release notes for every minor version (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 7. Maintenance & community, 89 out of 100, up to 1 more on the total

Why it scored 89: v5.5.0 tagged on 30 September 2026 (30). At least six releases between 15 July and 30 September, including v5.2.5, v5.3.0, v5.3.2, v5.4.0, v5.4.2 and v5.5.0 (20). The main branch had commits on 1 October 2026 and an issue-triage workflow runs, reply times not checked (15). The MCP App is in the official registry as io.github.tldraw/tldraw per the 30 September check (15). CI checks, a dependency dedupe workflow and a licence report (9).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- Checked 2026-10-01. The licence page says hobby licences send nothing and trial licences send only a hash of the key, with no user data. The SDK's LicenseManager sends the licence ID, SKU, environment and the full page URL (window.location.href) for trial, hobby-with-watermark and unlicensed production deployments, and the licence key page lists most of this. A full page URL can carry user data in its path or query (https://github.com/tldraw/tldraw/blob/main/packages/editor/src/lib/license/LicenseManager.ts, https://tldraw.dev/community/license).

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: tldraw.dev pages directly (our fetch was refused for rate limits); content read from the docs source in the GitHub repo
- Whether a canvas on the hosted MCP App can be read by anyone who knows its canvasId
- Rate-limit numbers on the hosted MCP App
- Whether default-branch CI is passing

## Weaknesses

- Source-available, and commercial prices aren't published
- `exec` runs model-written JavaScript on the canvas with no approval step
- Licence pings send the licence ID and full page URL for trial and hobby keys, more than the licence page says
- SECURITY.md lists only 3.x as supported while the current release is 5.5
- No server-side REST API, and the MCP App needs a host that renders MCP Apps

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Through the MCP App, call `search` on the Editor API spec first, then `exec` with JavaScript that calls `editor` methods
- Omit `canvasId` to start a blank canvas, and pass the returned `canvasId` to keep editing the same one
- Paste Mermaid into the canvas to get editable shapes instead of placing each shape by hand
- Return `editor.getCurrentPageShapes()` from `exec` to check what was drawn
- Read the release notes before a minor upgrade. Minor versions can break

## What the review panel asked for

- A headless export route
- Published commercial pricing
- list common exception texts

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.