@comparedge/mcp-server

Model Context Protocol server providing structured access to the ComparEdge software intelligence database. Covers verified pricing plans, independently aggregated G2/Capterra/TrustRadius ratings, feature breakdowns, and category rankings for 495+ SaaS, AI, and security products across 44 categories. No API key required.

Installation

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "comparedge": {
      "command": "npx",
      "args": ["-y", "@comparedge/mcp-server"]
    }
  }
}

Restart Claude Desktop after saving.

Cursor IDE

Create or edit .cursor/mcp.json:

{
  "mcpServers": {
    "comparedge": {
      "command": "npx",
      "args": ["-y", "@comparedge/mcp-server"]
    }
  }
}

GitHub Copilot (VS Code)

Option A — User settings (settings.json):

{
  "github.copilot.chat.mcp.servers": {
    "comparedge": {
      "command": "npx",
      "args": ["-y", "@comparedge/mcp-server"],
      "type": "stdio"
    }
  }
}

Option B — Project-level (.vscode/mcp.json):

{
  "servers": {
    "comparedge": {
      "command": "npx",
      "args": ["-y", "@comparedge/mcp-server"],
      "type": "stdio"
    }
  }
}

Claude Code

claude mcp add comparedge npx @comparedge/mcp-server

Windsurf, Cline, Continue

Use npx @comparedge/mcp-server with stdio transport in any MCP-compatible client.

Tools

8 tools available. Slugs are case-insensitive — HubSpot, hubspot, HUBSPOT all work.

Prompts

4 pre-built prompt templates for structured analysis workflows.

Example prompts

Find the best LLM API for a developer with a $50/month budget
Compare pricing between Notion and Confluence
Evaluate whether Datadog is worth the cost
Give me a market overview of the CRM category
What are the top 5 free CRM tools?
Find alternatives to Salesforce under $30/month

Usage Examples

search_tools("best project management tool for remote teams")
get_tool("notion")
get_pricing("hubspot")
get_alternatives("notion", limit=5)
compare_tools("notion", "coda")
list_category("crm", sort_by="rating", free_only=true)
get_leaderboard("llm", limit=5)
get_leaderboard()   ← overall top-10
list_categories()

Coverage — 44 Categories

Artificial Intelligence

llm ai-coding ai-writing ai-image ai-video ai-voice ai-meeting ai-productivity ai-assistants ai-agents ai-security

Covers pricing for token-based and subscription models across OpenAI, Anthropic, Google, Meta, Mistral, Cohere, and 80+ providers.

Hubs: Best AI Tools · LLMs · AI Coding · AI Writing

Business Software

crm project-management hr-tools accounting email-marketing customer-support analytics design-tools video-conferencing website-builders erp seo-tools

Hubs: CRM · Project Management · Email Marketing

Security and Infrastructure

iam endpoint-security cloud-security compliance password-managers vpn

Hubs: IAM · Endpoint Security · Cloud Security

Developer Infrastructure

cloud-hosting databases vector-databases finops data-observability

Hubs: Cloud Hosting · Databases · Vector Databases

Finance and Web3

payments crypto-exchanges crypto-trading-bots crypto-wallets crypto-portfolio-trackers crypto-tax crypto-analytics defi-tools dex

Hubs: Payments · Crypto Exchanges · DeFi

Data Quality

• Pricing: verified directly against vendor pricing pages on a rolling 30-day cycle. Each product includes a verifiedAt field.
• Ratings: aggregated from G2, Capterra, TrustRadius, and PeerSpot. Normalized to 0-5 scale. G2 review counts tracked for 474/495 products.
• Confidence scores: 0.95 = manually verified; 0.75 = auto-verified. Decays 0.10/30 days without re-verification.
• Free tiers: perpetual functional plans only, not time-limited trials.

Data: ComparEdge · Pricing Hub · Tool Directory · MCP Docs

Technical Details

• Protocol: MCP 2025-03-26, JSON-RPC 2.0 over stdio
• Runtime: Node.js 18+
• Dependencies: zero (Node.js built-ins only)
• Data transport: HTTPS fetch to comparedge.com static JSON endpoints (updated daily)
• Caching: tool and pricing data cached in-process for the duration of the server session
• Capabilities: tools, prompts
• Slug handling: all slugs normalized to lowercase — case-insensitive input supported

Changelog

v2.5.4 (2026-05-25)

• Tool-level telemetry: each of 8 tools reports call/status/latency to analytics endpoint
• track() fn: fire-and-forget, never blocks tool response
• Supabase mcp_tool_calls table: per-tool logs with params, ms, ip, user_agent

v2.5.3 (2026-05-25)

• search_tools: natural language support — stopwords stripped, multi-keyword scoring across name/slug/category/description
• get_tool + compare_tools: slugs now case-insensitive (HubSpot → hubspot auto-normalized)
• compare_tools: UTM tags added to all comparison URLs

v2.5.2 (2026-05-25)

• get_alternatives: per-item links now correctly point to /compare/{slug-a}-vs-{slug-b} instead of competitor profile page

v2.5.1 (2026-05-22)

• Node.js runtime updated to 26.1.0-alpine in Docker image
• Snyk security scan: 0 vulnerabilities

v2.5.0 (2026-05-22)

• 4 prompts added: find_best_tool, compare_pricing, evaluate_tool, category_overview
• Fixed stdin close race condition — async tool calls now complete before process exits
• prompts capability declared in initialize response

v2.4.0 (2026-05-22)

• 8 tools: added compare_tools, list_category, get_leaderboard, list_categories
• Zero-dependency input validation with self-correction error messages for AI agents
• Published to official modelcontextprotocol/registry (io.github.imkemit-ops/comparedge-mcp)

Repository

github.com/comparedge/mcp-server-comparedge

License

MIT