Magic Patterns vs Subframe

Magic Patterns 75

Hosted lines, for the MCP server and v3 API. A Statuspage at magicpatterns.statuspage.io with incident history, but its two components are Web Platform and Figma Plugin, with nothing for the API or MCP server (15 of 20). The history page and Atom feed post no incident since 9 January 2026, so the 90 days before this check are clean as far as the page covers (30). The page's /api/ path is disallowed by its robots.txt, so the JSON feeds weren't used. Rate limit published as 1,000 generations per 10 hours per key, with polling no faster than once every 60 seconds (15). No 429, Retry-After or backoff guidance and no idempotency keys. Design-system writes take a baseArtifactId and return 409 on a concurrent change, with re-read-and-retry advice (5 of 15). No SLA found below Enterprise's custom contracts (0). The MCP server and v3 API carry no beta label (10). 15 + 30 + 15 + 5 + 0 + 10 = 75.

Subframe 19

Read with the hosted lines, because the surface an agent uses is the MCP server at mcp.subframe.com. No status page was found on the site, the docs or the footer (0 of 20). With no page there is no incident history to read (5 of 30). No rate limit is published in the docs or the skill files (0 of 15). No 429 or backoff guidance was found. The docs do say a background job that stops reporting for about 10 minutes is returned as error, and the CLI retries a failed request once (4 of 15). No SLA is published, and the terms say the service is not warranted to be uninterrupted (0 of 10). The MCP server carries no beta label, though the CLI's push-component command is marked experimental (10 of 10).

Magic Patterns 84

The MCP server is closed, so its tool definitions weren't read. The v3 API has an OpenAPI 3.0.3 description, served as a block in each API reference page; the file llms.txt links at /docs/openapi-v3.yml answered with a not-found page (20 of 25). llms.txt and a Markdown copy of every docs page (10). The tools page sets out four workflows with call order, says when to poll and how often, splits read from write tools and warns that the user may edit at any time (17 of 20). Parameters are typed and marked required, the feedback category is an enum, and file writes take an array of name and content pairs (10 of 15). Request and response examples, an error table (400, 401, 402, 403) and 409 behaviour on writes (12 of 15). Versioned /v3 paths, a v2 deprecation notice and a dated product changelog (15). 20 + 10 + 17 + 10 + 12 + 15 = 84.

Subframe 54

The MCP server is closed and answers 401 without OAuth, so the tool definitions and their input schemas were not read. The docs list 46 tools by name and the public skill files give parameters for most of them (12 of 25, on that partial evidence). docs.subframe.com publishes llms.txt, a Markdown twin of each page and a docs MCP server (10 of 10). The design skill states for each tool what it is for and when another tool fits better, for example edit_page against design_page. The docs table gives one line a tool (15 of 20). Parameters are named in the skill (id, name or url, force, deleteChildPages, includeResolved), while the design tools take a free-text description (6 of 15). Ten example prompts and a troubleshooting section, with no list of error responses (7 of 15). The plugin and the CLI carry version numbers, with no changelog (4 of 15).

Magic Patterns 67

24 documented tools (15 for 11 to 30), plus 5 back for the read-only endpoint that registers only 12 (20 of 25). Message history pages by 10 and version history by 20 with skip, and file reads take a list of names (14 of 20). Documented status codes, a 402 that leaves reads working and non-blocking validationErrors on design-system writes (15 of 20). No idempotency keys; 409 conflict checks on design-system writes, and annotations couldn't be read on the closed server (10 of 20). create_design needs no parameter and makes a blank scaffold, but there is no official SDK (8 of 15). 20 + 14 + 15 + 10 + 8 = 67.

Subframe 47

46 documented tools is over 30 (5), with 3 added back because a Viewer account gets a read-only server. No toolsets or dynamic loading are documented, and the design skill adds about 55 KB when loaded (8 of 25). list_comments can be scoped to a page, a canvas or the project, screenshots can target one element or breakpoint and read_prototype_file returns one file. No paging is documented (10 of 20). wait_for_jobs reports running, done, error or not_found, delete tools refuse when the item is referenced, edit_page returns appliedCode and parser warnings, and the CLI prints a JSON error envelope (13 of 20). No idempotency keys. Tool annotations could not be read. Deletes need force when references exist, and version history can restore (8 of 20). projectId falls back to the first project and items are addressed by id, name or URL. No API SDK (8 of 15).

Magic Patterns 55

OAuth 2.1 with PKCE S256, dynamic registration and a device flow, issued by oauth.magicpatterns.com with WorkOS keys. Scopes are openid, email, profile and offline_access only, so a token carries the user's whole workspace role. API keys are shown once, with no scopes, expiry or rotation documented (22 of 30). The read-only endpoint doesn't register the 13 tools that create, change or publish. There is no confirmation step on the write endpoint (14 of 20). Tools return design code and chat history that collaborators wrote, and no prompt-injection guidance was found (3 of 15). x-mp-agent-name attributes changes in the UI, and audit logs come with Enterprise only (8 of 15). The security page describes encryption, logging and the Render and MongoDB Atlas setup and points to SOC 2 reports in a Trust Center drawn by script, which wasn't read. No security.txt, disclosure policy or bug bounty found (8 of 20). 22 + 14 + 3 + 8 + 8 = 55.

Subframe 44

The MCP server uses OAuth with dynamic client registration, PKCE and refresh tokens, and rejects static tokens. The authorisation server lists only identity scopes (openid, profile, email, phone, offline_access), so access follows the user's team role and not a scope. CLI tokens are shown once and can be deleted, with no scopes or expiry documented, and the generate_auth_token tool lets a connected agent mint one (22 of 30). Viewers get a read-only MCP server and read-only CLI access. Confirmation before deletes is guidance in the skill file and not a documented server control (12 of 20). list_comments, design documents and search_docs return text other people wrote, and no prompt-injection guidance was found (3 of 15). No audit log is documented. Version history records changes for 24 hours on Free and 7 days on Pro (3 of 15). No security.txt, disclosure policy, bug bounty or certification was found. The DPA's Annex 2 lists security measures and the docs cover Okta SSO (4 of 20).

Magic Patterns 23

No x402, MPP or L402. A 402 means out of credits, with no payment requirements (0). Seat prices are public and the on-demand rate is $0.02 a credit, but a generation's credit cost isn't known before it runs (typical 70 per generation per the docs), so between plan-only and per-unit (15 of 20). The MCP overview and the plans table say MCP and API-key access need a paid plan, while the v3 API page says free-tier accounts can call v3 up to their credits. Taking the stricter reading for the MCP surface, the free tier gives an agent no MCP access (5 of 20). A person signs in through the browser; dynamic client registration and the device flow still end with a human approval (3 of 20). 0 + 15 + 5 + 3 = 23.

Subframe 33

No machine payment protocol (0 of 40). Plan prices are public, Free at $0 and Pro at $20 an editor a month, with AI credits described only as limited and 6x and no per-call price (10 of 20). The Free plan has one project and includes MCP and CLI access, and the docs mention a card only at the Stripe checkout for Pro (20 of 20). A person has to sign in through a browser to approve OAuth. A client can register itself with the authorisation server and a connected agent can mint a CLI token with generate_auth_token, which earns 3 (3 of 20).

Magic Patterns 70

Changelog entry on 5 November 2026, the day before this check (30). Entries on twelve dates since 8 August, among them 23 September and 12 to 30 October (20). Closed service with a public changelog, email support within one to two business days, in-app chat and a Slack community. We didn't test them (12 of 15). Not in the official MCP registry (searches for magic-patterns and magicpatterns returned nothing of theirs). Official Claude connector, Cursor plugin and Codex plugin, and no SDKs (5 of 15). No public package or repository to check (3 of 10). 30 + 20 + 12 + 5 + 3 = 70.

Subframe 65

The Claude Code plugin reached 1.0.30 on 8 October 2026 and @subframe/cli 1.212.0 on 5 October 2026 (30 of 30). The plugin version changed ten times between 13 July and 8 October, and the CLI had three versions on 5 October after none since 24 June (20 of 20). The repository has 13 issues in all. Seven are open, among them a Codex MCP install error from 30 March 2026 and a Claude Code plugin install error from 13 February 2026 with one and two comments, and three with none. Pull requests merge most days. The Slack community was not read (10 of 25). A search of the official MCP registry for 'subframe' returned no entry (0 of 15). A commit of 12 August 2026 fixed dependency vulnerabilities. The CLI has unit and end-to-end tests in the repository, and the only public workflow publishes the package (5 of 10).