Webflow

# Fix list: Webflow

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/webflow, the October 2026 research run, assessed 8 October 2026. Grade B, 69.4 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 Webflow: 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. Payments & pricing, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No x402, MPP or L402 (0). Plan prices are public (Basic $15 and Premium $25 a month billed yearly, Team $2,500 a month on an annual contract), with Enterprise through sales and no per-call price (10). The Starter site plan is free and includes the CMS APIs at 60 requests a minute and the MCP server. The pricing page doesn't say whether signup needs a card (20, with that caveat). A person signs up in a browser and creates the token or approves the OAuth screen (0).

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.

## 2. Agent ergonomics, 72 out of 100, up to 4.6 more on the total

Why it scored 72: The MCP server documents 34 tools, each bundling several actions, which is the checklist's 5. We added 6 for `get_more_tools`, which loads specialised tools on request, and for tools labelled read or write (11 of 25). `limit` up to 100 and `offset` on lists, and custom-field filters and sorts on collection items with up to 10 filter terms (20). Errors carry a machine-readable code from a published list, and the MCP server returns `ModeForbidden` when a tool can't run in the current Designer mode (16). No idempotency keys on Data API writes. The open-source server sets `readOnlyHint` on its tools, and the hosted server has a dry run and an idempotency key for Webflow Cloud deploys only (10 of 20). Few required parameters, and official SDKs for JavaScript and Python (15).

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.

## 3. Security & auth, 74 out of 100, up to 4.6 more on the total

Why it scored 74: OAuth with read and write scope pairs per resource, or site tokens with the same scopes, at most 5 per site, which expire after 365 days without use. The MCP server uses OAuth with PKCE, dynamic client registration, refresh tokens and a revocation endpoint. Tokens travel in the Authorization header. We took 2 off because the MCP metadata also accepts the `plain` challenge method and site-token rotation isn't described (28). Scopes split read from write, the person picks sites or a workspace at authorisation, and the MCP server follows the user's role. A granted tool grants all its actions and no confirmation step for deletes was found (14). Agent Instructions stored on a site are given to every connected agent automatically, and the docs give no guidance on treating site content as untrusted beyond limiting authorised sites and reviewing generated drafts (5). Agent changes are recorded in the site activity log, which the pricing page lists on Team and Enterprise, with an Audit Logs API on Enterprise (9). A valid security.txt pointing to a Bugcrowd disclosure programme, SOC 2 Type II, ISO 27001, 27017, 27018 and 42001, and an annual penetration test report in the trust centre (18).

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.

## 4. Reliability, 79 out of 100, up to 4.2 more on the total

Why it scored 79: Graded on the Data API v2 and the hosted MCP server. Public status page at status.webflow.com with components for the Data API and the MCP server and a full incident history (20). Since 10 July 2026 the graded surface had four incidents, all marked minor (MCP tools missing from 9 to 16 July, outgoing webhooks on 28 July, 503 errors on the MCP server for about two hours on 25 August, MCP authorisation for client workspaces on 17 September). Two major incidents and one critical one in the same period were on hosted site pages and form emails, outside the graded surface. We read that as 15 of 30, between minor only and one major, because the July incident ran for a week. Rate limits published, 60 requests a minute on Starter and Basic and 120 on higher site plans (15). 429 carries Retry-After and the SDKs back off, but no idempotency keys were found for Data API writes (11). An SLA page dated 1 September 2026 exists and the pricing page lists enhanced SLAs on Enterprise. The terms sit in a PDF we didn't read (8 of 10). The /v2 namespace is described as production and the MCP server has a separate beta URL (10).

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.

## 5. Schema & documentation, 87 out of 100, up to 2.1 more on the total

Why it scored 87: Public OpenAPI 3.1 spec in webflow/openapi-spec with 140 operations for v2 and 139 for v2 beta, MIT. The hosted MCP server's v2.1 tool schemas aren't in the public repository, whose last commit is 10 April 2026 (23 of 25). A root llms.txt, section indexes and a Markdown copy of every page at the URL plus .md (10). The MCP data tools reference says for each tool whether it reads or writes, when to use it and what each action needs (16). The spec types parameters with enums and required fields. Item content in `fieldData` is an object checked against the collection's own schema (11). Request and response examples throughout, an error format with `code`, `message` and `details`, and an enumerated list of about 100 error codes (13). /v2 and /beta in the path and a dated changelog with a full index, though some changes land in place on the day (14).

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.

## 6. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: The newest developer changelog entry is dated 7 October 2026 (30). Sixteen dated entries since 10 July 2026, among them MCP v2.0.1 on 21 July and MCP v2.1 on 21 September (20). Public changelog, a support site and a community forum. We couldn't read the GitHub issue trackers, and the open-source MCP repository has had no commits since 10 April 2026 while the hosted server moved to v2.1 (8 of 15). The server is in the official MCP registry as com.webflow/mcp, and the JavaScript SDK (3.3.4, 16 March 2026) and Python SDK (2.0.0, 12 March 2026) are current (15). The JavaScript SDK repository has CI and a dependency patch merged on 6 October 2026. The MCP repository pins webflow-api 3.2.1 (7).

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.

## 7. Transparency & trust, 81 out of 100, up to 1.7 more on the total

Made of editorial 65, provenance 97.

Why it scored 81: Closed service with published Terms of Service and Developer Terms. The OpenAPI spec, both SDKs and the older open-source MCP server are MIT (17). Privacy policy effective 17 March 2025 and a DPA effective 15 November 2023 with deletion or return of personal data at termination. Retention is stated as for as long as necessary, with no periods (20). A dated v1 deprecation notice (31 March 2025) and a removal notice for the User Accounts APIs, but no written deprecation policy, and two breaking changes in the last 90 days landed on the day of their changelog entry (11). Sub-processor list updated 9 July 2026 with each company's country, nearly all in the USA. No choice of data region was found (17).

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):

- Terms of service: read, states 6 of the 7 things a reader expects, and has 1 clause that costs points (7.1 of 10)

## Deductions

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

- 7 October 2026. The Get Site Plan endpoint changed the `id` and `displayName` it returns for Starter sites and renamed some plans, in place. The changelog entry of the same date calls it a breaking change and no earlier notice was found. It is documented, so the deduction is small (https://developers.webflow.com/home/changelog/2026/10/7).

## 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: GitHub stars and open issues for webflow/mcp-server and webflow/js-webflow-api. The GitHub API refused us for its rate limit, so responsiveness on issues wasn't read
- unchecked: the SLA terms and uptime figure, which sit in a PDF linked from webflow.com/legal/sla
- unchecked: the hosted MCP server's v2.1 tool input schemas and annotations, which need a signed-in session. The public repository stops at v1.2 (10 April 2026)
- unchecked: whether CI passes on the default branches of the SDK and MCP repositories
- Not established whether signup for the free Starter plan needs a card. The pricing page lists it as free and doesn't say
- Not established which price belongs to monthly billing. The pricing text shows yearly-billed prices, and the page's price attributes suggest $25 a month for Basic and $39 for Premium billed monthly
- Not established whether the Starter plan can publish a single CMS item. The pricing table marks per-item publishing as absent on Starter and Basic
- Not established when MCP v2.0 shipped or how much notice its tool renames had. The migration guide lists renames across v1.3, v2.0, v2.0.1 and v2.1
- The rate limits page still uses older plan names (CMS, eCommerce, Business), while the pricing page lists Starter, Basic and Premium
- The lead's docs URL (data/docs/ai-tools) is an older page that still describes the Bridge App as required and remote authorisation as experimental. The current MCP docs are under developers.webflow.com/mcp

## Weaknesses

- The MCP server documents 34 tools (27 data, 3 Designer session, 4 utility), each with several actions, and a granted tool grants all its actions
- No idempotency keys on Data API writes in the reviewed documentation. Site publish is limited to one successful call a minute
- The MCP server can't create new localised CMS items. It reads and updates existing items in secondary locales
- The Starter plan allows 50 CMS items and 60 requests a minute. The site activity log that records agent changes is listed on Team ($2,500 a month) and Enterprise
- On 7 October 2026 Get Site Plan changed its `id` and `displayName` values in place, marked as breaking in the changelog entry of the same day

## 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.

- Send the token as `Authorization: Bearer` to https://api.webflow.com/v2. Ask for `cms:read` and `cms:write` only, plus `sites:write` if the task publishes
- Create or update items first, then call Publish Items or Publish Site. An item with `isDraft` true and a `lastPublished` date is live with unpublished changes
- Stay under 60 requests a minute on Starter and Basic and 120 on Premium, read X-RateLimit-Remaining, and wait for Retry-After on 429
- Page item lists with `limit` (maximum 100) and `offset`, and filter with `filter[<fieldSlug>][<operator>]`, up to 10 terms
- Upload an asset in two steps (create the asset with a file hash, then POST the bytes to the presigned URL). Write localised content with the item's `cmsLocaleId`

## 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.