browser-compat-mcp-server

Overview

Web platform compatibility for frontend work: per-browser support from MDN's @mdn/browser-compat-data, Baseline state and dates from web-features, and browserslist target resolution weighted by caniuse-lite usage figures. Every dataset ships inside the package, so there are no runtime network calls, no API key, no rate limit, and no upstream to be down — the same answers come back air-gapped. Runs as a stdio process or a local Streamable HTTP server.

Tools

Related MCP server: caniuse-mcp

Capability reference

browsercompat_list_reference tool

• One required topic: bcd_namespaces (12), bcd_browsers (17), browserslist_agents (19), baseline_states (4), groups (103), snapshots (11)

• Entries carry id, label, and detail, plus count, reported, bcd_browser, usage_percent, maps_from, or spec_url where the topic has them

• browserslist_agents gives each agent's browser-compat-data counterpart or null — the null ones can never be evaluated and always land in unchecked_targets

browsercompat_get_feature tool

• One feature string, 1–200 characters: a BCD key (css.selectors.has) or a web-features id (has); resolved_as echoes which one matched and how

• resolve: true accepts the search index's single unambiguous top hit; off by default, so a typo returns a miss rather than a confident answer about the wrong feature

• include_runtimes: true adds bun, deno, nodejs, and oculus rows to the 13 reported desktop and mobile browsers

• outcome is found | no_compat_data | miss — a miss is found: false with guidance, never an error

• A web-features id spanning more than one BCD key omits the per-key fields (support, status, limiting_browser, mdn_url, spec_urls) and returns compat_keys to re-call with

• Typed failure: invalid_feature_input (whitespace-only feature)

browsercompat_check_baseline tool

• Up to 20 BCD keys or web-features ids per call, 1–200 characters each; one result per entry, in input order

• Each result carries Baseline state and dates, limiting_browser, deprecated / experimental / discouraged, and usage_percent_excluded alongside the usage_source it is a share of

• usage_percent_excluded is absent — never zero — when the feature reaches no caniuse id

• all_widely_available answers the Baseline question alone: every entry resolved at widely, one miss forces it false, and deprecation does not enter it

• Typed failure: invalid_feature_input (a whitespace-only entry)

browsercompat_search_features tool

• query 1–100 characters, with optional namespace (one of the 12 BCD namespaces) and baseline (widely | newly | limited | not_mapped) filters; limit 1–50, default 10

• Every hit carries matched_on, the field that matched, so the six-tier ranking is inspectable rather than a score

• support_summary is one line across the seven Baseline core browsers, with — for unsupported and ? for unknown

• Zero hits are a successful empty result plus a notice naming which filter to drop; totalCount and truncated report matches beyond limit

• Typed failure: invalid_query (a query that normalizes to zero tokens)

browsercompat_compare_support tool

• Up to 20 features against a required targets browserslist query (defaults, > 0.5%, last 2 versions) — required so browserslist never falls back to config in the server's working directory

• verdict per feature: clears | fails | inconclusive | miss | ambiguous; failing_targets names each failing target with the verdict behind it (partial, prefixed, flagged, removed, unsupported, preview_only)

• unchecked_targets lists every target the server declined to judge, with no_bcd_browser | unknown_version | no_bcd_data; all_clear requires that list to be empty

• target_coverage_percent and unchecked_coverage_percent give the caniuse-derived traffic share of the evaluated and unevaluated tokens

• Typed failures: invalid_target_query, no_targets_resolved, invalid_feature_input

Data sources

CC BY 4.0 requires attribution wherever the caniuse data travels, so every response carrying a usage figure carries this string: Usage data from caniuse.com, © Can I Use contributors, CC BY 4.0. Figures are a share of the ~96.7% of global traffic caniuse tracks. Full license texts and notices are in THIRD_PARTY_NOTICES.md.

Features

Built on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.

Browser-compat-specific:

• All four datasets are bundled and loaded in process — no runtime network calls, no API key, no rate limit, and nothing to configure

• Baseline is read per browser-compat-data key from status.by_compat_key, never rolled up from the feature level, because keys under one feature legitimately disagree

• One shared resolver behind every tool: exact BCD key, then web-features id, then a moved redirect, and only under resolve: true the search index's single unambiguous hit

• Target versions are ordered by browser-compat-data's release index rather than parsed version strings, with the caniuse spellings normalized both directions (safari 16.0 ↔ 16, samsung 20 ↔ 20.0)

Agent-friendly output:

• Every response echoes data_version — the version of each bundled dataset behind the answer, since a pinned snapshot goes stale on exactly the newest features

• A verdict is never claimed for a browser that was not evaluated: unknown support moves the target into unchecked_targets and the feature to inconclusive

• Misses are results, not failures — found: false with guidance naming the next call, and typed error reasons carrying recovery hints for the input a caller has to fix

• Usage figures state the population they are a share of, and carry the caniuse attribution on every response that reports one

Getting started

Add the following to your MCP client configuration file:

{
  "mcpServers": {
    "browser-compat-mcp-server": {
      "type": "stdio",
      "command": "bunx",
      "args": ["@cyanheads/browser-compat-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "browser-compat-mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@cyanheads/browser-compat-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "browser-compat-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "MCP_TRANSPORT_TYPE=stdio",
        "ghcr.io/cyanheads/browser-compat-mcp-server:latest"
      ]
    }
  }
}

For Streamable HTTP, set the transport and start the server:

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 bun run start:http
# Server listens at http://localhost:3010/mcp

Prerequisites

• Bun v1.4.0 or higher (or Node.js v24+).

• No API keys, accounts, or network access required — every dataset ships with the package.

Installation

1. Clone the repository:

git clone https://github.com/cyanheads/browser-compat-mcp-server.git

2. Navigate into the directory:

cd browser-compat-mcp-server

3. Install dependencies:

bun install

4. Configure environment (optional):

cp .env.example .env
# edit .env if you want to override transport or logging defaults

Configuration

There are no server-specific environment variables: no API keys, no base URLs, and deliberately no browserslist configuration variable — the target query is always a tool input rather than ambient state. Only the framework transport settings apply.

See .env.example for the full list of optional framework overrides.

Running the server

Local development

# One-time build
bun run rebuild

# Run the built server
bun run start:stdio
# or
bun run start:http

bun run devcheck   # Lint, format, typecheck, security
bun run test       # Vitest test suite
bun run lint:mcp   # Validate MCP definitions against spec

Docker

docker build -t browser-compat-mcp-server .
docker run --rm -p 3010:3010 browser-compat-mcp-server

The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/browser-compat-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.

Project structure

The generated file tree is docs/tree.md.

Development guide

See CLAUDE.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic

• Use ctx.log for request-scoped logging, ctx.state for tenant-scoped storage

• Register new tools directly in src/index.ts

• Data integrity: read the bundled datasets as they are and preserve their uncertainty; never fabricate a support fact the data does not carry

Contributing

Issues are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

License

Apache-2.0 — see LICENSE for details.