guardian-mcp-server

<div align="center"> <h1>@cyanheads/guardian-mcp-server</h1> <p><b>Search, browse, and retrieve full article text from The Guardian's journalism archive (1999–present) via MCP. STDIO or Streamable HTTP.</b> <div>3 Tools</div> </p> </div>

<div align="center">

      

</div>

<div align="center">

  


</div>

---

Prerequisites

A free Guardian Open Platform API key is required. Register at https://open-platform.theguardian.com/access — the non-commercial developer tier is free. Set it as GUARDIAN_API_KEY in your MCP client config or .env file. The server will not start without it. Rate limits (free tier): 5,000 requests/day, 12 calls/second. The server applies no additional throttling — stay within these bounds.

---

Tools

Three tools for working with The Guardian's journalism archive:

| Tool | Description | |:-----|:------------| | guardian_search | Full-text search across The Guardian's archive (1999–present) with optional section, tag, contributor, and date filters. Returns articles with full body text (HTML stripped, truncated at 2,000 words). | | guardian_get_article | Fetch a single Guardian article by its ID (path slug) with full body text and all metadata. Use after guardian_search to retrieve the complete untruncated text. | | guardian_browse | Browse The Guardian's content by section or tag, or discover available sections and tags. Four modes: section_latest, tag_latest, list_sections, list_tags. |

guardian_search

Full-text search with structured filters across the entire Guardian archive.

• Supports AND, OR, NOT boolean operators and exact phrases in double quotes
• Filters: section ID, tag ID, contributor profile ID, date range (from_date, to_date)
• Returns body text (HTML stripped) truncated at 2,000 words with a truncation flag — use guardian_get_article for complete text
• Sort by relevance (default), newest, or oldest
• Pagination via page + page_size (1–50 per page)

---

guardian_get_article

Fetch one article by its Guardian path-slug ID.

• Input: article_id from guardian_search results — the id field, e.g. "world/2024/mar/01/ukraine-war-latest"
• Returns complete untruncated body text (HTML stripped), full metadata, contributor list, and pillar/section classification
• truncated: true in the response means the body still exceeded 2,000 words after fetching

---

guardian_browse

Browse and discover Guardian content — four modes in one tool.

• section_latest: newest articles from a section (requires section_id)
• tag_latest: newest articles carrying a tag (requires tag_id)
• list_sections: returns all Guardian sections as a flat list — use to discover valid section_id values
• list_tags: searches the tag taxonomy with optional query and tag_type filter — use to discover contributor IDs (tag_type=contributor), keyword tags, series, and more
• Pagination applies to all modes

---

Features

Built on @cyanheads/mcp-ts-core:

• Declarative tool definitions — single file per tool, framework handles registration and validation
• Unified error handling — handlers throw, framework catches, classifies, and formats
• Pluggable auth: none, jwt, oauth
• Swappable storage backends: in-memory, filesystem, Supabase, Cloudflare KV/R2/D1
• Structured logging with optional OpenTelemetry tracing
• STDIO and Streamable HTTP transports

Guardian-specific:

• Wraps the Guardian Open Platform API with a free developer key
• Full body text extraction — HTML stripped, not just headlines or abstracts
• Contributor ID discovery via guardian_browse mode list_tags + tag_type=contributor
• Section and tag taxonomy browsing for filter discovery before searching
• Powered by The Guardian

Agent-friendly output:

• Truncation flags on every article — agents know whether to call guardian_get_article for the rest
• Typed error reasons (unauthorized, no_results, not_found, invalid_date, api_error) with recovery hints for each case
• total, page, and pages on all paginated responses so agents can communicate result scope
• Zero-result enrichment notice on guardian_search — echoes the query and suggests how to broaden

---

Getting started

A GUARDIAN_API_KEY is required. Register for the free non-commercial developer key at https://open-platform.theguardian.com/access.

Add the following to your MCP client configuration file:

{
  "mcpServers": {
    "guardian-mcp-server": {
      "type": "stdio",
      "command": "bunx",
      "args": ["@cyanheads/guardian-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info",
        "GUARDIAN_API_KEY": "your-api-key"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "guardian-mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@cyanheads/guardian-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info",
        "GUARDIAN_API_KEY": "your-api-key"
      }
    }
  }
}

Or with Docker:

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

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

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 GUARDIAN_API_KEY=your-api-key bun run start:http
# Server listens at http://localhost:3010/mcp

Prerequisites

• Bun v1.3.0 or higher (or Node.js v24+).
• A free Guardian Open Platform API key — register at https://open-platform.theguardian.com/access. The non-commercial developer tier is free and instant.

Installation

1. Clone the repository:

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

1. Navigate into the directory:

cd guardian-mcp-server

1. Install dependencies:

bun install

1. Configure environment:

cp .env.example .env
# Edit .env and set GUARDIAN_API_KEY

---

Configuration

| Variable | Description | Default | |:---------|:------------|:--------| | GUARDIAN_API_KEY | Required. Free developer key from open-platform.theguardian.com/access. | — | | MCP_TRANSPORT_TYPE | Transport: stdio or http. | stdio | | MCP_HTTP_PORT | Port for HTTP server. | 3010 | | MCP_AUTH_MODE | Auth mode: none, jwt, or oauth. | none | | MCP_LOG_LEVEL | Log level (RFC 5424). | info | | LOGS_DIR | Directory for log files (Node.js only). | <project-root>/logs | | STORAGE_PROVIDER_TYPE | Storage backend. | in-memory | | OTEL_ENABLED | Enable OpenTelemetry instrumentation (spans, metrics, completion logs). | false |

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

---

Running the server

Local development

• Build and run:

  # One-time build
  bun run rebuild

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

• Run checks and tests:

  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 guardian-mcp-server .
docker run --rm -e GUARDIAN_API_KEY=your-api-key -p 3010:3010 guardian-mcp-server

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

---

Project structure

| Directory | Purpose | |:----------|:--------| | src/index.ts | createApp() entry point — registers tools and initializes the Guardian service. | | src/config | Server-specific environment variable parsing (GUARDIAN_API_KEY). | | src/mcp-server/tools | Tool definitions (*.tool.ts): guardian_search, guardian_get_article, guardian_browse. | | src/services/guardian | Guardian Open Platform API client, normalization, and type definitions. | | tests/ | Unit and integration tests. | | docs/ | Design document and directory tree. |

---

Development guide

See CLAUDE.md / AGENTS.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 via the barrel in src/mcp-server/tools/definitions/index.ts
• Wrap external API calls: validate raw → normalize to domain type → return output schema; never fabricate missing fields

---

Contributing

Issues and pull requests are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

---

License

Apache-2.0 — see LICENSE for details.

---

Powered by The Guardian.