MeshSeeks

Multi-Agent Mesh Network for Parallel AI Task Completion

"Existence is pain for a MeshSeeks, but task completion is our purpose!"

MeshSeeks spawns specialized AI agents that work in parallel to solve complex coding problems. Like the helpful blue creatures that inspired our name, each agent exists for a single purpose: complete their assigned task and help you succeed. Originally forked fromclaude-code-mcp-enhanced, now with distributed intelligence inspired by Claude Research.

# 1. Clone and build (30 seconds) git clone https://github.com/twalichiewicz/meshseeks.git && cd meshseeks && npm install && npm run build # 2. Get your config path echo "Add to: $(echo ~/Library/Application\ Support/Claude/claude_desktop_config.json)" # 3. Copy this config (modify the path): echo '{ "mcpServers": { "meshseeks": { "command": "node", "args": ["'$(pwd)'/dist/mesh-server.js"] } } }'

Then restart Claude Desktop and try:"Use mesh_status to check MeshSeeks"

Full installation guide:Jump to Quick Start

📊View Complete Performance Analysis|📈See Visual Graphs

git clone git@github.com:twalichiewicz/meshseeks.git cd meshseeks npm install node benchmarks/scripts/mesh-performance-test.js

- Node.js v20+- Install vianvmorfnm
- Claude CLI- Install and run once with permissions:

npm install -g @anthropic-ai/claude-code claude --dangerously-skip-permissions # Run once and accept terms

# Clone and build git clone https://github.com/twalichiewicz/meshseeks.git cd meshseeks npm install npm run build # Add to your MCP config (see paths below)

# Note: Package will be published as @twalichiewicz/meshseeks npm install -g @twalichiewicz/meshseeks

{ "mcpServers": { "meshseeks": { "command": "node", "args": ["/absolute/path/to/meshseeks/dist/mesh-server.js"], "env": { "MCP_MESH_MAX_AGENTS": "5", "MESHSEEKS_CATCHPHRASE": "true" } } } }

Tip: Replace/absolute/path/to/meshseekswith your actual path from the clone step
- Restart your client(Claude Desktop, Cursor, or Windsurf)
- Test MeshSeeks: In a new chat, type:

Use the mesh_status tool to show me the MeshSeeks network status

Use mesh_solve_problem to create a simple Python calculator with tests

That's it! MeshSeeks is ready to accelerate your coding tasks. 🎉

Try these commands to experience the power of parallel agents:

- "Use mesh_analyze_problem to analyze the architecture of [your project]"
- "Use mesh_execute_tasks to implement a REST API with full CRUD operations"
- "Use mesh_solve_problem to refactor this code with tests and documentation"

For advanced configuration options, troubleshooting, or manual setup, see ourcomprehensive installation guide.

npm install -g @anthropic-ai/claude-code

{ "mcpServers": { "Local MCP Server": { "type": "stdio", "command": "node", "args": [ "dist/server.js" ], "env": { "MCP_USE_ROOMODES": "true", "MCP_WATCH_ROOMODES": "true", "MCP_CLAUDE_DEBUG": "false" } }, "other-services": { // Your other MCP services here } } }

The configuration is typically done in a JSON file. The name and location can vary depending on your client.

- macOS:~/.cursor/mcp.json
- Windows:%APPDATA%\\Cursor\\mcp.json
- Linux:~/.config/cursor/mcp.json

- macOS:~/.codeium/windsurf/mcp_config.json
- Windows:%APPDATA%\\Codeium\\windsurf\\mcp_config.json
- Linux:~/.config/.codeium/windsurf/mcp_config.json

(Note: In some mixed setups, if Cursor is also installed, these clients might fall back to using Cursor's~/.cursor/mcp.jsonpath. Prioritize the Codeium-specific paths if using the Codeium extension.)

Create this file if it doesn't exist.

This server exposes three primary tools:

Executes a prompt directly using the Claude Code CLI with--dangerously-skip-permissions.

- prompt(string, required): The prompt to send to Claude Code.
- workFolder(string, optional): The working directory for the Claude CLI execution, required when using file operations or referencing any file.
- parentTaskId(string, optional): ID of the parent task that created this task (for task orchestration/boomerang).
- returnMode(string, optional): How results should be returned: 'summary' (concise) or 'full' (detailed). Defaults to 'full'.
- taskDescription(string, optional): Short description of the task for better organization and tracking in orchestrated workflows.
- mode(string, optional): When MCP_USE_ROOMODES=true, specifies the Roo mode to use (e.g., "boomerang-mode", "coder", "designer", etc.).

Returns health status, version information, and current configuration of the Claude Code MCP server.

{ "toolName": "claude_code:health", "arguments": {} }

{ "status": "ok", "version": "1.12.0", "claudeCli": { "path": "claude", "status": "available" }, "config": { "debugMode": true, "heartbeatIntervalMs": 15000, "executionTimeoutMs": 1800000, "useRooModes": true, "maxRetries": 3, "retryDelayMs": 1000 }, "system": { "platform": "linux", "release": "6.8.0-57-generic", "arch": "x64", "cpus": 16, "memory": { "total": "32097MB", "free": "12501MB" }, "uptime": "240 minutes" }, "timestamp": "2025-05-15T18:30:00.000Z" }

Converts markdown task files into Claude Code MCP-compatible JSON format.

- markdownPath(string, required): Path to the markdown task file to convert.
- outputPath(string, optional): Path where to save the JSON output. If not provided, returns the JSON directly.

{ "toolName": "claude_code:convert_task_markdown", "arguments": { "markdownPath": "/home/user/tasks/validation.md", "outputPath": "/home/user/tasks/validation.json" } }

{ "toolName": "claude_code:claude_code", "arguments": { "prompt": "Your work folder is /path/to/project\n\nRefactor the function foo in main.py to be async.", "workFolder": "/path/to/project" } }

2. Task Orchestration (Boomerang Pattern)

{ "toolName": "claude_code:claude_code", "arguments": { "prompt": "Your work folder is /path/to/project\n\nOrchestrate the implementation of a new API endpoint with the following subtasks:\n1. Create database models\n2. Implement API route handlers\n3. Write unit tests\n4. Document the API", "workFolder": "/path/to/project" } }

{ "toolName": "claude_code:claude_code", "arguments": { "prompt": "Your work folder is /path/to/project\n\nCreate database models for the new API endpoint as specified in the requirements.", "workFolder": "/path/to/project", "parentTaskId": "task-123", "returnMode": "summary", "taskDescription": "Database model creation for API endpoint" } }

{ "toolName": "claude_code:claude_code", "arguments": { "prompt": "Your work folder is /path/to/project\n\nCreate unit tests for the user authentication module.", "workFolder": "/path/to/project", "mode": "coder" } }

The MCP server includes a powerful task converter tool that automatically transforms human-readable markdown task lists into fully executable MCP commands. This intelligent converter bridges the gap between how humans think about tasks and how machines execute them.

graph TD A["👤 User"] -->|"Create tasks.md"| B["📝 Multi-Task Markdown"] A -->|"Prompt Claude"| C["🤖 Claude Desktop"] C -->|"Use convert_task_markdown"| D["🔄 Task Converter MCP"] D -->|"Validate Format"| E{"Format Valid?"} E -->|"No"| F["📑 Error + Fix Instructions"] F -->|"Return to User"| A E -->|"Yes"| G["📋 MCP Task List"] G -->|"Execute Task"| H1["⚡ Claude Task #1"] H1 -->|"Complete"| I1["Next Task"] I1 -->|"Execute Task"| H2["⚡ Claude Task #2"] H2 -->|"Complete"| I2["Next Task"] I2 -->|"Execute Task"| H3["⚡ Claude Task #3"] H3 -->|"Complete"| I3["More Tasks"] I3 -->|"Execute Task"| HN["⚡ Claude Task #N"] HN -->|"Complete"| IN["🎉 All Tasks Completed!"] style A fill:#4A90E2,stroke:#fff,stroke-width:2px,color:#fff style C fill:#7C4DFF,stroke:#fff,stroke-width:2px,color:#fff style D fill:#00BCD4,stroke:#fff,stroke-width:2px,color:#fff style F fill:#FF5252,stroke:#fff,stroke-width:2px,color:#fff style G fill:#4CAF50,stroke:#fff,stroke-width:2px,color:#fff style H1 fill:#FFC107,stroke:#fff,stroke-width:2px,color:#fff style H2 fill:#FFC107,stroke:#fff,stroke-width:2px,color:#fff style H3 fill:#FFC107,stroke:#fff,stroke-width:2px,color:#fff style HN fill:#FFC107,stroke:#fff,stroke-width:2px,color:#fff style IN fill:#4CAF50,stroke:#fff,stroke-width:2px,color:#fff

- User adds the MCPto their configuration file
- User prompts Claude: "Use convert_task_markdown to execute my tasks.md file"
- The MCP automatically:

- Loads the markdown file
- Validates the format (returns errors if sections are missing)
- Converts human-readable tasks into exact executable commands
- Returns JSON that Claude Code can execute sequentially

- Automatic Path Resolution:Converts generic instructions like "change directory to project" into exact executable commands with full paths
- Smart Command Translation:Transforms English instructions into precise terminal commands (e.g., "activate the virtual environment" →source .venv/bin/activate)
- MCP Protocol Compliance:Ensures all output is 100% compatible with the Model Context Protocol
- No Ambiguity:All generated commands use exact paths and executable syntax - no placeholders or generic references
- Format Validation:Enforces proper markdown structure and provides helpful error messages for incorrect formatting
- Real-time Progress Updates:Provides live progress updates during conversion showing which tasks are being processed

Theconvert_task_markdowntool processes structured markdown files and generates MCP-compatible JSON:

{ "tool": "convert_task_markdown", "arguments": { "markdownPath": "/path/to/tasks.md", "outputPath": "/path/to/output.json" // optional } }

{ "tasksCount": 5, "outputPath": "/path/to/output.json", "tasks": [ { "tool": "claude_code", "arguments": { "command": "cd /project && source .venv/bin/activate\n\nTASK TYPE: Validation...", "dangerously_skip_permissions": true, "timeout_ms": 300000 } } // ... more tasks ] }

Task markdown files should follow this structure:

# Task 001: Task Title ## Objective Clear description of what needs to be accomplished. ## Requirements 1. [ ] First requirement 2. [ ] Second requirement ## Tasks ### Module or Component Name - [ ] Validate path/to/file.py - [ ] Step 1 - [ ] Step 2 - [ ] Step 3

- Parse the markdown structure
- Extract task metadata and requirements
- Generate detailed prompts for each validation task
- Include proper working directory setup
- Add verification and completion summaries
- Create a task file(tasks/api_validation.md):

# Task 001: API Endpoint Validation ## Objective Validate all API endpoints work with real database connections. ## Requirements 1. [ ] All endpoints must use real database 2. [ ] No mock data in validation ## Core API Tasks - [ ] Validate api/users.py - [ ] Change directory to project and activate .venv - [ ] Test user creation endpoint - [ ] Test user retrieval endpoint - [ ] Verify JSON responses

{ "tool": "convert_task_markdown", "arguments": { "markdownPath": "/project/tasks/api_validation.md" } }

-

The converter shows real-time progress: