Skip to main content
Glama
Spider-Chat

spiderchat-mcp-server

Official
by Spider-Chat
README.md
# Spider Chat MCP Server

Access your Spider Chat flows, memories, and facts from Claude Desktop.

## Installation

### 1. Install
```bash
npm install
npm run build
```

### 2. Get Your API Token
1. Go to https://nonlinearaichat-production.up.railway.app/
2. Dashboard → API Tokens
3. Generate New Token
4. Copy it

### 3. Configure Claude Desktop

**macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`  
**Windows:** `%APPDATA%\Claude\claude_desktop_config.json`

Add this (replace the path and token):

```json
{
  "mcpServers": {
    "spiderchat": {
      "command": "node",
      "args": ["/ABSOLUTE/PATH/TO/THIS/FOLDER/dist/index.js"],
      "env": {
        "SPIDERCHAT_API_TOKEN": "your_token_here",
        "SPIDERCHAT_API_URL": "https://nonlinearaichat-production.up.railway.app"
      }
    }
  }
}
```

**Get the absolute path:**
```bash
cd mcp-server && pwd
# Use the output + /dist/index.js
```

### 4. Restart Claude Desktop

Quit completely and reopen.

## Usage

Ask Claude:
- `"Search my Spider Chat flows for machine learning"`
- `"Show me my most important memories"`
- `"List all my flow groups"`
- `"Find everything about Python in my Spider Chat"`

## Troubleshooting

**Claude doesn't see it:**
- Use absolute path (not relative)
- Make sure you ran `npm run build`
- Completely quit and restart Claude

**Authentication failed:**
- Check your token is correct
- Generate a new token from dashboard

**Connection error:**
- No trailing slash in URL: `https://nonlinearaichat-production.up.railway.app`

## Features

- 🔍 Search flows, memories, and facts
- 📁 Browse flow groups
- 🔒 Read-only access
- ⚡ Fast semantic search

## Requirements

- Node.js 18+
- Claude Desktop
- Spider Chat account

## License

MIT

TDQS

B3.4/5.0

Scored across 8 tools

Disambiguation4/5

Most tools have clearly distinct purposes (search_flows, search_memories, search_facts target different resources; unified_search is a combined search). However, get_flow_group and list_flows_in_group both relate to group contents and could be confused, though descriptions differentiate the group details from the list of flows.

Naming Consistency4/5

Tool names mostly follow a verb_noun pattern with snake_case (search_flows, get_flow_nodes, list_flow_groups). The exception is unified_search, which uses an adjective instead of a verb, creating a minor inconsistency.

Tool Count5/5

8 tools is well within the ideal range for a focused MCP server. Each tool covers a distinct need (searching, listing, retrieving), and there is no redundancy or excessive fragmentation.

Completeness3/5

The tool surface covers search and browsing well, with flows, memories, facts, and group navigation. However, there is no direct retrieval by ID for memories or facts, no way to list all flows without going through groups, and no create/update/delete operations, which limits full lifecycle coverage if such functionality is expected.

Maintenance

ActivityInactive
ResponsivenessNo issues