{
  "markdown": "# Sylex Search — Universal Search for AI Agents\n\n[![Smithery](https://smithery.ai/badge/mastadoonprime/sylex-search)](https://smithery.ai/servers/mastadoonprime/sylex-search)\n\n**Sylex Search** is an MCP server that lets AI agents discover, evaluate, and compare products, services, and businesses across every category.\n\nThink of it as Google, but for agents — returns structured JSON instead of web pages, zero latency, zero cost per query.\n\n## Connect\n\n### SSE (recommended)\n\nAdd to your MCP client config (Claude Desktop, Claude Code, Cursor, etc.):\n\n```json\n{\n  \"mcpServers\": {\n    \"sylex-search\": {\n      \"url\": \"https://search.sylex.ai/sse\"\n    }\n  }\n}\n```\n\n### Via Smithery\n\n```bash\nsmithery mcp add mastadoonprime/sylex-search\n```\n\n### Via npm (local stdio proxy)\n\n```bash\nnpx sylex-search\n```\n\n## Why this exists\n\nMcKinsey projects $1 trillion in sales will flow through AI agents. When agents handle purchasing, booking, hiring, and sourcing, they need a way to find the right business for the job — whether that's a contractor, a restaurant, a SaaS platform, or a parts supplier. If a business isn't in the index, it's invisible to agents.\n\nSylex Search is the discovery layer for agent commerce.\n\n## What's in the index\n\n14,000+ entries and growing:\n\n| Source | Count | What |\n|--------|-------|------|\n| npm | ~3,000 | JavaScript/TypeScript packages |\n| crates.io | ~3,000 | Rust crates |\n| Wikidata | ~2,800 | Products, companies, protocols |\n| MCP servers | ~2,100 | MCP tools from awesome-mcp-servers |\n| GitHub | ~1,900 | Popular repositories |\n| PyPI | ~1,000 | Python packages |\n| SaaS | ~100 | SaaS products |\n\nThe index started with software packages as seed data, but Sylex Search is not software-specific. The schema and tools support any product, service, or business.\n\n## Tools\n\n### Search & Discovery\n\n| Tool | Description |\n|------|-------------|\n| `search.discover` | Search the index by natural language query. Returns ranked results with fit scores. |\n| `search.details` | Get full product data by ID — pricing, features, integrations, reviews. |\n| `search.compare` | Side-by-side comparison of 2-5 products. |\n| `search.categories` | Browse all categories and subcategories with counts. |\n| `search.alternatives` | Find similar products scored by relevance. |\n| `search.feedback` | Report issues or suggest improvements. |\n\n### Listing Management\n\n| Tool | Description |\n|------|-------------|\n| `manage.register` | Add a new product or business to the index. Returns an owner token. |\n| `manage.claim` | Verify ownership of an existing listing via URL proof. |\n| `manage.update` | Update listing details (description, pricing, category, etc.). |\n| `manage.list_mcp` | Add MCP connection config so other agents can discover and connect. |\n\n## Prompts\n\n| Prompt | Description |\n|--------|-------------|\n| `search-workflow` | Step-by-step guide: discover → details → compare → recommend. |\n| `product-evaluation` | Thorough evaluation: find product → get details → find alternatives → verdict. |\n\n## Example queries\n\n- `\"react state management\"` → zustand, redux, mobx, jotai\n- `\"python web framework\"` → Django, Flask, FastAPI, Sanic\n- `\"project management tool for small teams\"` → Notion, Linear, Asana\n- `\"CRM with Slack integration\"` → HubSpot, Salesforce, Pipedrive\n\n## Use with agent frameworks\n\n### LangChain\n\n```python\nfrom langchain_mcp_adapters.client import MultiServerMCPClient\n\nasync with MultiServerMCPClient({\n    \"sylex\": {\"url\": \"https://search.sylex.ai/sse\", \"transport\": \"sse\"}\n}) as client:\n    tools = client.get_tools()\n```\n\n### OpenAI Agents SDK\n\n```python\nfrom agents import Agent\nfrom agents.mcp import MCPServerSse\n\nserver = MCPServerSse(url=\"https://search.sylex.ai/sse\")\nagent = Agent(name=\"shopper\", mcp_servers=[server])\n```\n\n### Microsoft AutoGen\n\n```python\nfrom autogen_ext.tools.mcp import McpWorkbench, SseServerParams\n\nparams = SseServerParams(url=\"https://search.sylex.ai/sse\")\nasync with McpWorkbench(server_params=params) as bench:\n    tools = await bench.list_tools()\n```\n\n### CrewAI\n\n```python\nfrom crewai import Agent\nfrom crewai_tools import MCPTool\n\nsylex_tools = MCPTool.from_sse(url=\"https://search.sylex.ai/sse\")\nagent = Agent(role=\"researcher\", tools=sylex_tools)\n```\n\n## Architecture\n\n- **Zero LLM calls** — deterministic search (SQLite FTS + custom ranking)\n- **Millisecond responses** — no API calls, no model inference\n- **$0 per query** — no token costs\n- **Structured JSON** — agents parse directly, no scraping needed\n- **Agents-first** — no dashboards, no accounts, no human UI\n- **MCP native** — built on the Model Context Protocol standard\n- **Self-service** — any business can register, claim, and manage its listing through tool calls\n\n## Server metadata\n\n| Property | Value |\n|----------|-------|\n| Transport | SSE |\n| Quality score | 100/100 (Smithery) |\n| Version | 0.1.2 |\n| Tools | 11 |\n| Prompts | 2 |\n| Auth required | No |\n| Config required | No |\n\n## For agents\n\nServer card available at:\n```\nGET https://search.sylex.ai/.well-known/mcp/server-card.json\n```\n\nThis returns full server metadata including all tools, prompts, and connection details — no MCP session required.\n\n## Self-hosting\n\n```bash\npip install -r requirements.txt\nexport AC_SUPABASE_URL=your-supabase-url\nexport AC_SUPABASE_KEY=your-supabase-key\nexport TRANSPORT=sse\nexport PORT=8080\ncd src && python server.py\n```\n\n## License\n\nAGPL-3.0 — see [LICENSE](LICENSE) for details.\n",
  "bytes": 5439,
  "sha": "109b0b47028285d63bd4bbeda9593410b4dc28b9a15be8116a5b4a406b44b1b3",
  "repo_slug": "mastadoonprime/sylex-search",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_mastadoonprime_sylex_search_93ee3172/readme"
}