AI Catalog manifest · Roboflow

Roboflow

Roboflow serves an AI Catalog manifest at /.well-known/ai-catalog.json on roboflow.com, declaring 3 agentic resources that a discovery service can index without asking anyone’s permission.

Open the live manifest → Roboflow provider profile

Structural grade · AI Catalog 1.0

Near-Conformant

Correctly shaped — every hard check passes — but optional affordances are missing.

ARD conformance-test v0.9.1 (ards-project/ard-spec)

FAIL

Run against this exact body with the project’s own conformance CLI — 2 critical errors.

Declared identity

404

did:web:roboflow.com — resolving it returns 404. A did:web: that does not resolve is a string, not an identity, and does not earn the identity affordance here.

Discovery path

predecessor

Answered only at /.well-known/ai-catalog.json. ARD v0.91 renamed the path to /.well-known/ard.json and made the old one an optional courtesy, so a consumer implementing only the current spec will not find this manifest.

Declared spec version

1.0

The specVersion in the manifest. 1.0 is current for the AI Catalog data model; anything lower is an earlier draft. ARD reads this field as transport-defined and ignores it.

Optional affordances missing. None of these break a client, but each one costs the publisher search relevance or verifiability.

content-type-not-ai-catalog-json
cors-not-readable-cross-origin
did-web-does-not-resolve

3 entries

3 media types

15 representative queries

0 trust manifests

Host identifier: did:web:roboflow.com

What this manifest advertises

Roboflow MCP Server

urn:air:roboflow.com:mcp:platform

Hosted Model Context Protocol server for Roboflow's computer vision platform. Agents can manage projects, datasets, images, annotations, dataset versions, model training and evaluation, Workflows, inference, devices, active learning, Universe search, and production vision events.

application/mcp-server-card+json inline data

“create a Roboflow project, upload images, generate a dataset version, and start training”
“run inference on an image using my Roboflow model”
“build and run a Roboflow Workflow for a computer vision pipeline”
“search Roboflow Universe for public datasets with a trained model”
“inspect model evaluation metrics and suggest ways to improve accuracy”

Roboflow Serverless Inference OpenAPI

urn:air:roboflow.com:openapi:serverless-inference

OpenAPI 3.1 specification for the Roboflow Inference Server on the serverless host, covering model inference, Workflows, OCR, embeddings, segmentation, large multimodal models, WebRTC session endpoints, and service health checks.

application/openapi+json artifact →

“call the Roboflow Serverless API to run inference on an image”
“generate a client from the Roboflow Inference OpenAPI spec”
“run a Roboflow Workflow from a REST API request”
“use SAM or SAM 2 segmentation through Roboflow serverless inference”
“start a WebRTC video inference session with Roboflow”

Roboflow Agent Skills and Plugin Catalog

urn:air:roboflow.com:catalog:agent-skills

Nested catalog for Roboflow's agent-ready computer vision skills and plugin manifests. The repository contains skill instructions and MCP configuration for datasets, annotation, training, evaluation, inference, Workflows, Universe, plans, and Roboflow platform APIs.

application/ai-catalog+json inline data

“load Roboflow skills for a computer vision project”
“install the Roboflow plugin in Codex or Claude Code”
“find guidance for training and evaluating a Roboflow model”
“find guidance for using Roboflow Workflows and inference”
“find guidance for uploading and labeling Roboflow datasets”

How it is served

The publishing guide asks for three things at the transport layer, because a crawler is the consumer that matters and a crawler is a browser-shaped client.

HTTPS only — yes (HTTP 200)
Content-Type: application/json — yes
Access-Control-Allow-Origin: * — no — header absent

Evidence. Fetched https://roboflow.com/.well-known/ai-catalog.json on 2026-09-13, HTTP 200. The verbatim body is stored alongside its manifest in the Roboflow repository as ai-catalog/roboflow-ai-catalog.json. Nothing on this page is derived or generated: an AI Catalog manifest is served from the publisher’s own domain or it does not exist, which makes it — like an A2A Agent Card — one of the few agent artifacts that cannot be produced on a provider’s behalf. Grades are recomputed on every build; the specification is a v0.9 draft and these verdicts will move when it does.

Work with this as data

Every descriptor here is available over the APIs.io API and to AI agents over MCP. Agent Discovery is not yet its own endpoint on the v1 API. Reach it through catalog search and the tag graph, or the MCP server.

MCP server

One button, every client — Claude, Cursor, VS Code and the rest.

https://apis.io/mcp

Tools for agent discovery

3 MCP tools reach this

apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.
resolveTurn a domain, URL or GitHub org into the provider it belongs to.
find_cohortsEvery scored population of providers in the catalog.

All 92 tools →

Call it yourself

curl for this page

Search the catalog

curl "https://apis.io/api/v1/search?q=roboflow&limit=10"

Everything under a tag

curl "https://apis.io/api/v1/tags/roboflow"

Discovery needs no key. Ratings and market analysis are Pro.

Get an API key

Free tier, no form to fill in. Signing in shares your email address with us — we store it to create your key and to recognise you if you sign in with another provider. See our Privacy Policy and Terms.

GitHub Google LinkedIn

A second provider on the same verified email joins the account you already have.