claude mcp search

Search MCP server for Claude Desktop & Claude Code.

Add searchable cloud storage to Claude in one line of config. Sprigr gives your agent persistent memory that survives between conversations: full-text search, structured data, instant retrieval.

Start free See how it works
sprigr · new workflow listening
Example run · 41 actions hover the gate to approve it yourself
the claude mcp search back office

Persistent storage across every conversation. Full-text search over all stored records. Structured JSON data with filterable attributes. The agent self-manages its own knowledge base. Instant retrieval, no re-pasting needed.

You focus on the work. Sprigr runs the paperwork.

what it handles
  • Persistent memory
  • Full-text + semantic
  • One-line setup
  • Multi-index
what it runs

Built for Claude, designed for agents

Every feature is optimized for the way MCP tools work inside Claude Desktop and Claude Code.

  • One-line setup

    A single JSON block in your MCP config file. The server runs through npx, so there is nothing to install globally, no Docker containers, no Python environments. Paste, save, restart. Done.

  • Smart import, deduplicated memory

    The smart_import tool checks each new record against what Claude already stored before writing. Duplicates merge, stale facts are superseded with history kept, and new records insert. Agent memory stays clean without manual housekeeping.

  • Scoped API keys

    Restrict a key to specific indexes, or mint a secured key that locks a row-level filter into every query. One agent, one user, one slice of the data.

  • Full-text & semantic search

    Typo-tolerant keyword search plus optional hybrid semantic search. Claude can find records by meaning. A query for "frustrated customer" finds complaints even if those exact words aren't in the record.

  • Faceted filtering

    Tag records with attributes like category, project, date, or priority. Claude can filter search results by any combination of facets to find precisely the right data.

  • Persistent across conversations

    Data lives in your Sprigr account, not in Claude's context window. Close a conversation, start a new one tomorrow. Every record is still there and searchable.

  • Multi-index support

    Create separate indexes for different projects or data types. Research notes in one index, code references in another, contacts in a third, all searchable from the same MCP server.

how it starts

Set up Claude MCP search in 30 seconds

Three steps from zero to searchable agent memory.

  1. 01

    Add the config

    Paste the Sprigr MCP server entry into your claude_desktop_config.json. One JSON block, no dependencies to install, no build step.

  2. 02

    Add your API key

    Sign up at platform.sprigr.com, copy your API key from the admin panel, and set it as SPRIGR_API_KEY in the config. Every future session connects automatically.

  3. 03

    Store & search

    Claude uses the import_objects tool to push records and the search tool to query them. Notes, code snippets, research, anything you want the agent to remember.

Open claude_desktop_config.json and add "sprigr": { "command": "npx", "args": ["-y", "@sprigr/mcp-server"], "env": { "SPRIGR_API_KEY": "your-api-key" } } inside mcpServers. Restart Claude Desktop and 20 Sprigr tools are available in every conversation.

See the MCP docs →
questions

Questions

How do I install the Sprigr MCP server for Claude?

Add the Sprigr MCP server entry to your claude_desktop_config.json file. On macOS this is at ~/Library/Application Support/Claude/claude_desktop_config.json. On Windows it is at %APPDATA%\Claude\claude_desktop_config.json. Add the "sprigr" block inside the "mcpServers" object, save the file, and restart Claude Desktop. The server runs locally via npx, so there is nothing to build or host.

Does Claude remember my API key between conversations?

Yes. The API key is stored in your claude_desktop_config.json file, so it persists across all Claude Desktop sessions. You only need to sign up once, and the key is available in every future conversation automatically.

What can I store in Sprigr through Claude?

Any structured JSON data. Meeting notes, code snippets, research summaries, contact lists, project metadata, bookmarks, reading notes, anything you want Claude to remember and search across sessions. Each record is a JSON object with any fields you define.

Is the search fast enough for agent use?

Yes. Sprigr's search engine returns results in milliseconds for typical indexes. Claude can make multiple search calls in a single response without noticeable latency. Full-text queries, filtered queries, and semantic queries are all optimized for agent workflows.

Does it work with Claude Code as well as Claude Desktop?

Yes. In Claude Code run claude mcp add sprigr -e SPRIGR_API_KEY=your-api-key -- npx -y @sprigr/mcp-server, or add the same mcpServers block to a .mcp.json in your project. You get the same 20 tools in your terminal-based workflows, which suits storing coding patterns, config snippets, and project notes.

claude mcp search

Give Claude persistent memory.

Paid plans from $49 a month.

Start free Explore the platform