# AgentIndex x402 MCP Discovery

> AgentIndex x402 MCP Discovery is a paid API for AI agents from x402.agentindex.world, paid per call via x402, $0.001/call, status unknown (last checked 2026-10-02).

Semantically searches a snapshot of 10,000+ MCP servers to find the best-matching server for a plain-language capability need

## Facts

- Endpoint: GET https://x402.agentindex.world/discover?utm_source=zero.xyz
- Price: $0.001/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-02
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agentindex-x402-mcp-discovery-cd2270b3
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_qAiKxzOjfVSeFJ2yquFNf

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability agentindex-x402-mcp-discovery-cd2270b3
```

Example prompt: Search the MCP registry for servers that can read a PDF and return clean markdown — show me the top 3 results with at least 0.4 similarity.

## When to prefer this

Choose this endpoint when an AI agent needs to programmatically discover which MCP server best fits a particular capability need, rather than maintaining a hardcoded list of tools. It is especially useful for meta-agent architectures that dynamically select tools at runtime, or for agent builders exploring what MCP tooling exists for a given domain. Prefer it over manual registry browsing or hardcoded tool lists when the capability requirement is expressed in natural language.

## Known failure modes

- Query string too vague — returns low-relevance or zero results below the min_similarity threshold
- min_similarity set too high — filters out all candidates and returns an empty results array
- Snapshot may be days old — newly published MCP servers may not appear
- Truncated description in crawled schema may cause unexpected input validation errors
- Payment failure via x402/USDC on Base blocks the call entirely

## How this service works

Find MCP servers matching a need, ranked by semantic similarity over a curated snapshot of 10101 MCP servers. Returns name, endpoint, description, source registry and a 0-1 relevance per match. Paid POST: up to 25 matches. Try GET /discover/sample.

## Output

Returns a JSON object containing the matched query string, the number of servers above the similarity threshold, and an ordered list of the top matching MCP servers — each with name, description, registry source, endpoint URL (if known), and cosine relevance score — plus snapshot metadata (date taken, total rows in snapshot).

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "q": {
   "type": "string",
   "description": "The need to match, in plain language - e.g. \"read a PDF and give me markdown\" or \"persistent knowledge graph\". Matched against a curated snapshot of MCP servers (official registry + carnet) by semantic similarity."
  },
  "max_results": {
   "type": "integer",
   "maximum": 25,
   "minimum": 1,
   "description": "Maximum number of servers to return (default 5)."
  },
  "min_similarity": {
   "type": "number",
   "maximum": 1,
   "minimum": 0,
   "description": "Minimum cosine similarity to include a result (default 0.30). Raise it for fewer, closer matches."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "q": "read a PDF and give me markdown",
  "matches": 3,
  "results": [
   {
    "url": "https://example.org/mcp/",
    "name": "PDF Extract MCP",
    "registry": "registre-mcp",
    "relevance": 0.7211,
    "description": "Extract a public PDF into clean markdown text, plus metadata and a real token count.",
    "observed_at": "2026-09-02T19:07:49.374328Z"
   },
   {
    "url": null,
    "name": "Docling Server",
    "registry": "annuaire",
    "relevance": 0.6487,
    "description": "Document conversion to markdown and structured text for LLM agents.",
    "observed_at": "2026-09-10T12:00:00Z"
   },
   {
    "url": "https://example.com/mcp/",
    "name": "Srclight",
    "registry": "registre-mcp",
    "relevance": 0.4102,
    "description": "Deep code indexing for AI agents. FTS5 + embeddings + call graphs. Fully local.",
    "observed_at": "2026-09-15T08:30:00Z"
   }
  ],
  "snapshot_date": "2026-09-18",
  "snapshot_rows": 10101,
  "min_similarity": 0.3
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agentindex-x402-mcp-discovery-cd2270b3/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from x402.agentindex.world](https://www.zero.xyz/host/x402.agentindex.world/llms.txt)