# Deep Agentic Core MCP

> Deep Agentic Core MCP is an MCP server for ai models and llm tooling. One MCP server for profiling, chaos-testing and supervising agentic workflows. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Platforms:** CLI, API
- **Repository:** https://github.com/deepagentlabs/mcp-server
- **MCP Registry name:** io.github.DeepAgentLabs/deep-agentic-core-mcp
- **Categories:** [AI models and LLM tooling](https://ai-powered.com/categories/ai-and-llm), [developer tools](https://ai-powered.com/categories/developer-tools)
- **Listing updated:** 2026-11-18

## About

The DeepAgentLabs libraries each solve one problem for agent builders: AgenticLens inspects and diagnoses workflow runs, Agentic Chaos injects faults to test resilience, and Agentic Sidecar covers supervision readiness. This package puts them behind one MCP identity so a client connects once instead of installing a server per library. It sits above the AI Operations Workflow Specification, the shared data contract.

Tools are namespaced. The README names lens.analyze_workflow (recommendations with provenance blocks that cite artifact digests and span ids), lens.diagnose_run (failure diagnoses and a faulty-step ranking), chaos.list_faults, chaos.run_experiment, chaos.export_aios_run, spec.validate_artifact (version-aware validation against the spec's rules) and core.compatibility and core.verify for checking installed pieces.

A warning matters here: chaos.run_experiment executes real workspace code, so the server is meant for trusted local stdio use. Script globs can allow or deny what it may run, and hosted HTTP deployments reject chaos execution. The README is open that version 1.0.0 does not certify production readiness. A hosted service at mcp.deepagentlabs.io also exists.

## Features

- **Workflow analysis:** lens.analyze_workflow returns recommendations with provenance down to span ids.
- **Run diagnosis:** lens.diagnose_run reports failure diagnoses and the likely faulty step.
- **Fault injection:** chaos.run_experiment runs controlled chaos experiments such as stream cuts, locally.
- **Spec validation:** spec.validate_artifact checks artifacts against the schema set their spec version names.
- **Script allow and deny globs:** Environment variables narrow which workspace scripts chaos runs may execute.
- **Compatibility reports:** core.compatibility and core.verify report installed packages and probe adapters.

## Good for

- Find the faulty step in a failed agent run
- Test how an agent workflow behaves when a stream is cut
- Validate an exported workflow against the AI Operations spec
- Check which sibling packages are installed and compatible

## Install

### Claude Code (Terminal)

```sh
claude mcp add deep-agentic-core-mcp -- uvx deep-agentic-core-mcp
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "deep-agentic-core-mcp": {
      "command": "uvx",
      "args": [
        "deep-agentic-core-mcp"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "deep-agentic-core-mcp": {
      "command": "uvx",
      "args": [
        "deep-agentic-core-mcp"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "deep-agentic-core-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "deep-agentic-core-mcp"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "deep-agentic-core-mcp": {
      "command": "uvx",
      "args": [
        "deep-agentic-core-mcp"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "deep-agentic-core-mcp": {
      "command": "uvx",
      "args": [
        "deep-agentic-core-mcp"
      ]
    }
  }
}
```

## Tools

- `core.compatibility`: Reports installed distribution observations.
- `core.verify`: Probes adapters.
- `chaos.list_faults`: Lists available fault types.
- `chaos.run_experiment`: Runs a chaos experiment against a local target script.
- `chaos.export_aios_run`: Converts a saved or session report to an AI Operations run.
- `lens.analyze_workflow`: Analyzes a workflow and returns recommendations with provenance.
- `lens.diagnose_run`: Diagnoses a trace run and ranks candidate faulty steps.
- `spec.validate_artifact`: Validates an artifact against its spec version's schema.

## FAQ

### Is it safe to run?

chaos.run_experiment executes real code, so the README says to use stdio on trusted local machines; hosted HTTP deployments reject chaos execution.

### Does 1.0.0 mean production ready?

No. The README says package versions do not certify AIOS conformance or production deployment readiness.

### Is there a hosted version?

Yes, at https://mcp.deepagentlabs.io/mcp, where everything except chaos.run_experiment behaves the same.

### How do I install it?

From PyPI as deep-agentic-core-mcp, typically as a stdio server.

### How do I add Deep Agentic Core MCP to Claude Code?

Run this command in your terminal: claude mcp add deep-agentic-core-mcp -- uvx deep-agentic-core-mcp

### What are the best alternatives to Deep Agentic Core MCP?

The closest alternatives to Deep Agentic Core MCP are REA, Context7, Chrome DevTools MCP, Codebase Memory and GitHub.

## Where this comes from

- Description: written from the project's README.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Context7](https://ai-powered.com/mcp/context7): Context7 is a freemium MCP server for developer tools.
- [Chrome DevTools MCP](https://ai-powered.com/mcp/chrome-devtools-mcp-2): Chrome DevTools MCP is an open-source MCP server for browser automation.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [awesome-mcp-servers](https://ai-powered.com/mcp/awesome-mcp-servers-2): awesome-mcp-servers is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/deep-agentic-core-mcp
