Skip to main content
Glama
sofya-mcp

sofya-mcp

Official
by sofya-co

sofya-mcp

Listed on mcpservers.org

MCP server for Sofya - web search, fetch, extract, and deep research for AI agents, exposed over the Model Context Protocol.

Use it either way:

  • Hosted (no install): connect your client to https://mcp.sofya.co/mcp (streamable HTTP). Sign in with OAuth, or send your API key as a Bearer token. This is the server listed in the MCP registry as co.sofya/sofya.

  • Local: build this repo and run it over stdio. It is a thin wrapper around the Sofya REST API with the four core tools.

Eligible GitHub accounts (a year or older) get free monthly credits; everyone else pays as they go.

Tools

Tool

What it does

Cost

search

Search the web and get extracted page content (not just snippets). Supports news topic, domain filters, freshness, and an optional AI-synthesized answer.

1-2 credits (+2 for include_answer)

fetch

Fetch up to 10 URLs as clean markdown. Also handles PDF, DOCX, XLSX, PPTX, EPUB, and more.

1 credit / URL

extract

Fetch a page and pull specific information from it using AI.

10 credits

research

Decompose a question into sub-queries, read many sources in parallel, and synthesize a cited report.

50 credits

get_balance

Show the account's remaining credits. Hosted server only.

free

buy_credits

Return a payment link for a credit top-up. It does not charge anything. Hosted server only.

free

search, fetch, extract, research and get_balance are read-only.

Related MCP server: mcp-web-search

Hosted server (VS Code / Copilot and others)

Point your client at the hosted endpoint. Clients with OAuth support (Claude, VS Code, Cursor, Cline and others) only need the URL and will ask you to sign in. To use an API key instead, get one at sofya.co (keys look like sofya_live_...) and send it as a Bearer token:

VS Code / GitHub Copilot (.vscode/mcp.json):

{
  "servers": {
    "sofya": {
      "type": "http",
      "url": "https://mcp.sofya.co/mcp",
      "headers": { "Authorization": "Bearer sofya_live_..." }
    }
  }
}

Claude Code, as a plugin (signs in with OAuth; plugin files in plugins/sofya):

/plugin install sofya --marketplace sofya-co/sofya-mcp

Claude Code, with an API key:

claude mcp add --transport http sofya https://mcp.sofya.co/mcp --header "Authorization: Bearer sofya_live_..."

Cursor (~/.cursor/mcp.json):

{
  "mcpServers": {
    "sofya": {
      "url": "https://mcp.sofya.co/mcp",
      "headers": { "Authorization": "Bearer sofya_live_..." }
    }
  }
}

Cline:

cline mcp install sofya --transport http https://mcp.sofya.co/mcp

Or in cline_mcp_settings.json:

{
  "mcpServers": {
    "sofya": {
      "type": "streamableHttp",
      "url": "https://mcp.sofya.co/mcp",
      "headers": { "Authorization": "Bearer sofya_live_..." }
    }
  }
}

Then ask your agent things like "search for the latest Node.js LTS release notes", "fetch this PDF and summarize it", or "research how the EU AI Act treats open-weight models". Full API docs: sofya.co/docs.

Agent skills

Two skills in skills (the same files ship inside the Claude plugin at plugins/sofya/skills) teach an agent when and how to use Sofya: sofya-search (search, fetch, extract) and sofya-research (cited reports). They use the MCP tools when they are connected, and otherwise call the REST API with SOFYA_API_KEY. Install them into Claude Code, Cursor, Codex, Gemini CLI, Copilot and other agents with:

npx skills add sofya-co/sofya-mcp

Claude Code plugin

The plugin bundles the hosted server and both skills:

/plugin marketplace add sofya-co/sofya-mcp
/plugin install sofya@sofya

Codex

The repo root is also a Codex plugin (.codex-plugin/plugin.json) with the hosted server and both skills.

Kiro

This repo is also a Kiro power (plugin.json, mcp.json and skills at the root). In Kiro, open Powers, choose Add Custom Power, then Import power from GitHub and enter https://github.com/sofya-co/sofya-mcp. Kiro opens a Sofya sign-in the first time a tool runs.

Local server

Prerequisites

  • Node.js 18+

  • A Sofya API key - get one at sofya.co (keys look like sofya_live_...)

Build

git clone https://github.com/sofya-co/sofya-mcp.git
cd sofya-mcp
npm install          # also builds to build/index.js

Check it starts:

SOFYA_API_KEY=sofya_live_... node build/index.js

In the examples below, replace /path/to/sofya-mcp with the folder you cloned.

Configuration

Claude Code

claude mcp add sofya --env SOFYA_API_KEY=sofya_live_... -- node /path/to/sofya-mcp/build/index.js

Claude Desktop / Cursor / Windsurf / VS Code

Add this to your MCP config (e.g. claude_desktop_config.json, or .cursor/mcp.json):

{
  "mcpServers": {
    "sofya": {
      "command": "node",
      "args": ["/path/to/sofya-mcp/build/index.js"],
      "env": {
        "SOFYA_API_KEY": "sofya_live_..."
      }
    }
  }
}

Codex

Add it with the Codex CLI:

codex mcp add sofya --env SOFYA_API_KEY=sofya_live_... -- node /path/to/sofya-mcp/build/index.js

Or edit ~/.codex/config.toml directly. Note Codex puts environment variables in a nested [mcp_servers.<name>.env] table:

[mcp_servers.sofya]
command = "node"
args = ["/path/to/sofya-mcp/build/index.js"]

[mcp_servers.sofya.env]
SOFYA_API_KEY = "sofya_live_..."

Then run /mcp inside Codex to confirm the server is connected.

Environment variables

Variable

Required

Default

Description

SOFYA_API_KEY

yes

-

Your Sofya API key (sofya_live_...; older ay_live_... keys still work).

SOFYA_BASE_URL

no

https://sofya.co

Override the Sofya API base URL.

CLI flags --api-key and --base-url take precedence over the env vars. Run node build/index.js --list-tools to print the tool schemas, or --help for options.

Development

npm run build        # compile TypeScript to build/
npm run watch        # recompile on change
npm run inspector    # debug with @modelcontextprotocol/inspector

The server is a single file, src/index.ts. Tool definitions mirror Sofya's hosted MCP server exactly (same names, parameters, and defaults).

Privacy and support

The hosted server receives the tool arguments (queries, URLs, prompts) to answer each request and keeps them only in server logs, up to 30 days. Privacy policy · Terms

Support: support@sofya.co or sofya.co/contact.

License

MIT - see LICENSE.

Available Tools

4 tools
extractAInspect

Fetch a webpage and extract specific information using AI. Use this when you need structured data from a page (e.g. pricing, specs, contact info) rather than the raw content. Costs 5 credits.

Returns: content (the extracted text), url, credits_used, credits_remaining, usage (token counts).

Inputs: url (required), prompt (required).

fetchAInspect

Fetch one or more URLs and return their content as clean markdown. Use this to read articles, documentation, blog posts, or any page where you need the complete text, not just a snippet from search. Also supports PDF, DOCX, and other document formats. Costs 1 credit per URL. Max 10 URLs per request. Failed URLs are not charged.

Set include_raw_html=true to also get the raw HTML source in each result. Useful for inspecting embedded URLs, data attributes, iframes, or script tags that are stripped during markdown conversion. Returns null for non-HTML content (PDF, DOCX, etc.). Same cost.

Returns: results (array of {title, url, content, raw_html, published_time, success, error}), credits_used, credits_remaining.

Inputs: urls (required), include_raw_html.

researchAInspect

Perform comprehensive research on a topic. Decomposes your query into sub-queries, searches and reads multiple sources in parallel, then synthesizes a structured report with citations. Best for open-ended or comparative questions that need coverage from many angles. For simple factual lookups, use search instead (optionally with include_answer=true for cheap synthesis). Costs 25 credits.

Returns: query, report (structured markdown with citations), sources (array of {title, url, fetched}), sub_queries (the decomposed queries), credits_used, credits_remaining, usage (token counts).

Inputs: query (required), topic, freshness, max_sources.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.5/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: extract for structured data from a single page, fetch for raw content, research for synthesized reports, and search for web searches. No overlap that would confuse an agent.

    Naming Consistency5/5

    All tool names are single verbs (extract, fetch, research, search) following a consistent pattern. No mixed conventions or unclear naming.

    Tool Count5/5

    Four tools is well-scoped for a server focused on web information retrieval and synthesis. Each tool earns its place without being too few or too many.

    Completeness5/5

    The tool set covers all major operations: fetching raw content, extracting structured data, searching the web, and performing deep research. No obvious gaps for the stated purpose.

    Related MCP Connectors

    Related MCP Servers