Connect to Exodata's exoplanet MCP server

Exodata provides hosted Model Context Protocol access to NASA Exoplanet Archive records. Connect an AI agent to inspect catalog schemas, run read-only SQL, explore curated insights, and export a planet or stellar host as JSON or CSV. Exodata is an independent service, not operated by NASA.

Connection URL:

• Hosted: https://exodata.space/mcp

Transport: Streamable HTTP in stateless JSON response mode. Simple request/response clients do not need to manage MCP session IDs.

Quick setup for Codex, Claude Code, and OpenCode

Choose your client; no local MCP server or proxy installation is needed.

Codex CLI:

codex mcp add exodata --url https://exodata.space/mcp

Claude Code, shared project configuration:

claude mcp add --scope project --transport http exodata https://exodata.space/mcp

OpenCode, project configuration:

opencode mcp add exodata --url https://exodata.space/mcp

These commands save client configuration. Restart or reconnect your client as needed, then ask it to call Exodata's health tool to check the connection. The server exposes the six tools listed below. Configuration details and other HTTP clients are covered under Connecting an Agent.

Supported data and refresh behavior

Which NASA Exoplanet Archive tables can I query?

Exodata loads prepared exports of NASA's stellarhosts and ps tables. Query them as stellarhosts and exoplanets respectively. Inspect available fields, descriptions, units, and types with describe_catalog; see the NASA PS column definitions for source field documentation.

Does each row represent a different planet?

No. exoplanets contains PS reference records: multiple rows can describe the same planet using different references. COUNT(*) counts reference records; use COUNT(DISTINCT pl_name) to count distinct planet names. Host joins can also return multiple matching records. Check the schema and choose the rows appropriate for your question before interpreting a comparison or ranking.

Is the data queried live from NASA?

No. Queries run against the prepared catalog loaded at Exodata server startup, not against NASA's live service. New exports require conversion and a server restart before they appear. Exodata does not promise a daily refresh schedule. Record fields such as rowupdate describe source rows, not the service's last refresh date.

Can I use SQL and download results?

Use query_catalog for a single read-only SQL SELECT, including supported joins and aggregations. This is Exodata's SQL interface, not an upstream NASA ADQL endpoint. Query responses are bounded by the limits below; download_detail exports one named planet or host as JSON or CSV, rather than an unlimited query result. Browse the catalog, stellar hosts, or REST API guide for related access paths.

Tools

query_catalog uses the same SQL validation and execution path as /rest/query.

Limits

The MCP limits are stricter on purpose: agents pull rows into context windows, so capping responses prevents one query from saturating the conversation.

Queries that fail validation, exceed the 30-second timeout, or target an unregistered table return MCP errors with a descriptive message.

Recommended Agent Flow

1. Call describe_catalog for the table and columns relevant to the task.
2. Use the returned metadata to write a SQL SELECT.
3. Call query_catalog with the SQL and optional limit.

Example describe_catalog arguments:

{
  "table": "exoplanets",
  "columns": ["pl_name", "hostname", "pl_rade", "pl_eqt"]
}

Example query_catalog arguments:

Ask for reference records with planet radii between 0.8 and 1.2 Earth radii. Inspect the returned records rather than treating this radius filter as a habitability assessment; one planet may occur more than once.

{
  "sql": "SELECT pl_name, hostname, pl_rade FROM exoplanets WHERE pl_rade BETWEEN 0.8 AND 1.2 ORDER BY pl_rade",
  "limit": 25
}

Join example:

{
  "sql": "SELECT s.hostname, s.st_teff, e.pl_name, e.pl_rade FROM stellarhosts s JOIN exoplanets e ON s.hostname = e.hostname ORDER BY e.pl_rade LIMIT 10"
}

Aggregate example:

This counts reference records by discovery method, not distinct planets.

{
  "sql": "SELECT discoverymethod, COUNT(*) AS count FROM exoplanets GROUP BY discoverymethod ORDER BY count DESC",
  "limit": 20
}

Example download_detail arguments:

{
  "entity": "exoplanet",
  "name": "Kepler-22 b",
  "format": "json"
}

download_detail returns filename, mime_type, content, and the matching browser download url. Supported entities are stellarhost and exoplanet; supported formats are json and csv.

Connecting an Agent

The hosted endpoint speaks Streamable HTTP, so any MCP client with native HTTP transport support can connect without a local proxy. Each client uses a slightly different config file name and key.

Claude Code

Run from the project root with explicit project scope:

claude mcp add --scope project --transport http exodata https://exodata.space/mcp

That writes .mcp.json. Replace --scope project with --scope user for global configuration in ~/.claude.json. Omitting --scope defaults to a private local-project entry in ~/.claude.json, not .mcp.json. See Claude Code's MCP guide.

Project configuration:

{
  "mcpServers": {
    "exodata": {
      "type": "http",
      "url": "https://exodata.space/mcp"
    }
  }
}

Crush

Add this to crushrc in the project root or ~/.config/crush/crushrc, following the Crush configuration guide:

mcp add exodata --type http --url https://exodata.space/mcp

Existing JSON configurations (.crush.json, crush.json, or ~/.config/crush/crush.json) remain supported as a legacy format:

{
  "$schema": "https://charm.land/crush.json",
  "mcp": {
    "exodata": {
      "type": "http",
      "url": "https://exodata.space/mcp"
    }
  }
}

OpenCode

The quick-setup command writes to project configuration. Add --global to write to global configuration instead. Alternatively, add a remote entry to opencode.json in the project root or ~/.config/opencode/opencode.json, following the OpenCode MCP guide:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "exodata": {
      "type": "remote",
      "url": "https://exodata.space/mcp",
      "enabled": true
    }
  }
}

Codex CLI

Codex supports Streamable HTTP natively. The quick-setup command above adds the hosted URL to ~/.codex/config.toml. You can instead add the entry directly, following the official Codex MCP guide:

[mcp_servers.exodata]
url = "https://exodata.space/mcp"

After configuration, the agent should see six tools: health, list_insights, run_insight, describe_catalog, query_catalog, and download_detail. Call health first to confirm the connection.