Sub-Agent-MCP

KO | EN

gitlite — search

#python #hacktoberfest #typescript #go #deep-learning #macos #nodejs #windows #awesome #kubernetes #security #mcp #openai #open-source

(★ 13)

MCP for spawning sub-agents everywhere

Sub-Agent-MCP Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Sub-Agent MCP

[![CI](https://github.com/systemgroupnet/Sub-Agent-MCP/actions/workflows/ci.yml/badge.svg)](https://github.com/systemgroupnet/Sub-Agent-MCP/actions/workflows/ci.yml)
[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://raw.githubusercontent.com/systemgroupnet/Sub-Agent-MCP/main/LICENSE)
[![Python](https://img.shields.io/badge/python-%3E%3D3.10-blue.svg)](https://www.python.org/downloads/)
[![Container](https://img.shields.io/badge/ghcr.io-sub--agent--mcp-2496ED?logo=docker&logoColor=white)](https://github.com/systemgroupnet/Sub-Agent-MCP/pkgs/container/sub-agent-mcp)

Production-ready Python MCP server for **LLM delegation and sub-agent orchestration**. A parent LLM (for example, Cursor’s agent) connects to this server and delegates work by calling a **tool named after each agent** defined in YAML (for example, `researcher`).

Each sub-agent is defined in YAML with its own LLM, system prompt, and optional downstream MCP tool servers.

## Table of contents

- [What is this?](#what-is-this)
- [Why use it?](#why-use-it)
- [Features](#features)
- [Prerequisites](#prerequisites)
- [Quick start](#quick-start)
- [Verify installation](#verify-installation)
- [Connect Cursor](#connect-cursor)
- [How it works](#how-it-works)
- [Agent configuration](#agent-configuration)
- [MCP tools reference](#mcp-tools-reference)
- [Environment variables](#environment-variables)
- [Project layout](#project-layout)
- [Development](#development)
- [Docker image and releases](#docker-image-and-releases)
- [Troubleshooting](#troubleshooting)
- [License](#license)

## What is this?

Sub-Agent MCP sits between a **parent LLM** and one or more **specialized sub-agents**:

1. The parent connects to this server over **Streamable HTTP** at `/mcp`.
2. At startup, each agent in `agents.yaml` is registered as an MCP tool named by its `id`.
3. The parent calls that tool with a `prompt`; the server runs the sub-agent and returns the final response.

Each sub-agent is a [LangChain](https://github.com/langchain-ai/langchain) agent with its own OpenAI-compatible LLM, system prompt, and optional connections to other MCP servers (filesystem, search, your own tools, and so on). Tool access can be restricted per agent with an allowlist.

```mermaid
sequenceDiagram
    participant Parent as Parent_LLM
    participant SAMCP as Sub_Agent_MCP
    participant Agent as LangChain_sub_agent
    participant LLM as OpenAI_compatible_LLM
    participant Tools as Downstream_MCP_servers

Parent->>SAMCP: researcher(prompt)
    SAMCP->>Agent: build agent + downstream tools
    Agent->>LLM: reasoning loop
    Agent->>Tools: MCP tool calls
    Tools-->>Agent: tool results
    LLM-->>Agent: final answer
    Agent-->>SAMCP: response text
    SAMCP-->>Parent: {"response": "..."}
```

This is different from giving one agent every tool in the workspace: the parent stays lightweight, roles stay explicit, and each sub-agent only sees the MCP servers and tools you configure for it.

## Why use it?

- **Delegation without context bloat** — The parent calls agent tools directly; it does not need every downstream tool schema in its own context.
- **Per-role configuration** — Different `id`s can use different models, prompts, MCP servers, and tool allowlists.
- **Production-oriented** — Pydantic-validated YAML, structured logging, Docker health checks, CI, and GHCR images on release tags.
- **OpenAI-compatible providers** — Point `llm.base_uri` at OpenAI, Azure, Ollama, LM Studio, or any compatible API.

## Features

**Transport**

- Streamable HTTP only (`streamable-http`); no stdio or legacy SSE.

**Configuration**

- YAML agent definitions with strict Pydantic validation.
- Environment substitution: `${VAR}` and `${VAR:-default}`.

**Runtime**

- LangChain 1.x `create_agent` with OpenAI-compatible chat models.
- Per-agent MCP connections via `langchain-mcp-adapters`, with optional `server.tool` allowlists.

**Security**

- Agent tool descriptions never expose API keys.

**Operations**

- Structured logging ([structlog](https://www.structlog.org/)).
- Docker image with health check; GitHub Actions CI.
- Container images published to GHCR on version tags (`v0.x.y`).

**MCP tools exposed by this server**

Each agent in `agents.yaml` is registered as a tool named by its `id` (for example, `researcher`). Each tool accepts a single `prompt` argument and runs that sub-agent.

## Prerequisites

| Requirement                                                  | Notes                                                       |
| ------------------------------------------------------------ | ----------------------------------------------------------- |
| Python 3.10+                                                 | CI uses 3.12; `requires-python >= 3.10` in `pyproject.toml` |
| API key for your LLM provider                                | Default example uses `OPENAI_API_KEY`                       |
| [uv](https://github.com/astral-sh/uv) (recommended) or `pip` | Matches CI install path                                     |
| Docker + Compose (optional)                                  | Recommended for first run; includes mock MCP tool servers   |

## Quick start

Choose one path. **Docker Compose is the fastest way** to run the full demo stack (server + mock tool servers).

### Path A — Docker Compose (recommended)

```bash
export OPENAI_API_KEY=sk-...
docker compose up --build
```

| Service             | Port | Endpoint                    |
| ------------------- | ---- | --------------------------- |
| Sub-Agent MCP       | 8000 | `http://localhost:8000/mcp` |
| Mock filesystem MCP | 8001 | `http://localhost:8001/mcp` |
| Mock search MCP     | 8002 | `http://localhost:8002/mcp` |

The bundled [config/agents.yaml](config/agents.yaml) uses **Docker network hostnames** (`filesystem-mcp`, `search-mcp`), which resolve correctly inside Compose.

### Path B — Local Python

```bash
uv sync --dev
# or: pip install -e ".[dev]"

cp config/agents.example.yaml config/agents.yaml   # if you don't have config/agents.yaml yet
export OPENAI_API_KEY=sk-...

python -m sub_agent_mcp.main
# or: uv run sub-agent-mcp
```

Server listens at `http://0.0.0.0:8000/mcp` (reachable as `http://localhost:8000/mcp` from your machine).

**Important:** The example config points MCP servers at Docker service names. For local Python you must either run the mock servers and use `localhost` URLs (see table below), or run only the mocks via Compose:

```bash
# Terminal 1: mock tool servers only
docker compose up filesystem-mcp search-mcp

# Terminal 2: sub-agent server (after editing agents.yaml URLs to localhost)
export OPENAI_API_KEY=sk-...
python -m sub_agent_mcp.main
```

#### Local vs Docker MCP URLs

| Environment                 | filesystem MCP URL               | search MCP URL               |
| --------------------------- | -------------------------------- | ---------------------------- |
| Docker Compose network      | `http://filesystem-mcp:8001/mcp` | `http://search-mcp:8002/mcp` |
| Host machine / local Python | `http://localhost:8001/mcp`      | `http://localhost:8002/mcp`  |

### Path C — Prebuilt image (GHCR)

Images are published on **git tags** matching `v*` (for example `v0.1.2`), not on every push to `main`.

```bash
docker run -p 8000:8000 \
  -e OPENAI_API_KEY=sk-... \
  -v "$(pwd)/config/agents.yaml:/app/config/agents.yaml:ro" \
  ghcr.io/systemgroupnet/sub-agent-mcp:latest
```

Mount your own `agents.yaml` and ensure MCP `url` values are reachable from inside the container (use host networking, service names, or `host.docker.internal` as appropriate).

## Verify installation

**OpenAPI document** (generated from registered MCP tools):

```bash
curl -s http://localhost:8000/mcp/openapi.json | head
```

**Open WebUI (OpenAPI tool server)** — Register as **OpenAPI** (not MCP):

| Field | Value |
|-------|--------|
| **URL** | `http://localhost:8000/mcp` (or `http://mcp.example.com/sub-agent/mcp` behind a proxy) |
| **OpenAPI path** | `openapi.json` (default) |

Open WebUI appends tool paths from the spec (for example `/tools/researcher`) to that URL:

```bash
curl -s -X POST http://localhost:8000/mcp/tools/researcher \
  -H 'Content-Type: application/json' \
  -d '{"prompt":"Summarize what tools you have access to."}'
```

**Docker health** — The image health check probes `http://127.0.0.1:8000/mcp`.

**Functional check** — After connecting Cursor (below), ask the parent agent to call the `researcher` tool with a short prompt. A successful run returns `{ "response": "..." }`.

## Connect Cursor

1. Open **Cursor Settings → MCP** (or edit your MCP config JSON).
2. Add the server:

```json
{
  "mcpServers": {
    "sub-agent-mcp": {
      "url": "http://localhost:8000/mcp"
    }
  }
}
```

3. Ensure the Sub-Agent MCP process is running and Cursor can reach `localhost:8000`. On WSL or Docker Desktop, confirm port forwarding if the server runs in a VM or container.
4. Reload MCP tools. You should see one tool per agent (for example, **`researcher`**).

**Example delegation prompt for the parent agent:**

> Call the `researcher` tool with prompt: "Summarize what tools you have access to."

**Other MCP clients** — Any client that supports **Streamable HTTP** can connect to `http://<host>:8000/mcp`. Refer to your client’s MCP documentation for URL-based server configuration; this server does not use stdio transport.

## How it works

```mermaid
flowchart LR
    subgraph parent [Parent]
        Cursor[Cursor agent]
    end
    subgraph subagentmcp [Sub_Agent_MCP]
        ToolsMCP[agent tools e.g. researcher]
        Builder[Agent builder]
    end
    subgraph subagent [Sub_agent runtime]
        LC[LangChain create_agent]
        LLM[OpenAI_compatible LLM]
    end
    subgraph downstream [Per_agent MCP servers]
        FS[filesystem MCP]
        SRCH[search MCP]
    end

Cursor -->|Streamable HTTP /mcp| ToolsMCP
    ToolsMCP --> Builder
    Builder --> LC
    LC --> LLM
    LC --> FS
    LC --> SRCH
```

**Agent tool flow**

1. Load and validate [config/agents.yaml](config/agents.yaml) (or `AGENTS_CONFIG_PATH`) at startup.
2. Register one MCP tool per agent, named by `id`.
3. When a tool is called, build an OpenAI-compatible chat model from that agent’s `llm.*`.
4. Connect to the agent’s `mcp_servers`, discover tools, apply `tool_allowlist` if set.
5. Run the LangChain agent loop (bounded by `AGENT_RECURSION_LIMIT`).
6. Return the final assistant message as `{ "response": "..." }`, or `{ "error": "..." }` on failure.

## Agent configuration

Agents are loaded from `config/agents.yaml` at startup. Override the path with `AGENTS_CONFIG_PATH`.

```yaml
agents:
  - id: researcher
    title: Research Agent
    description: "Agent specialized in research tasks"
    llm:
      base_uri: https://api.openai.com/v1
      api_key: ${OPENAI_API_KEY}
      model_id: gpt-4.1-mini
    system_prompt: |
      You are a helpful research assistant.
    mcp_servers:
      - name: filesystem
        transport: streamable_http
        url: http://filesystem-mcp:8001/mcp
        headers: {}
      - name: search
        transport: streamable_http
        url: http://search-mcp:8002/mcp
    tool_allowlist:
      - filesystem.read_file
      - search.web_search
```

Copy [config/agents.example.yaml](config/agents.example.yaml) as a starting point.

### Schema reference

| Field                   | Description                                                                                  |
| ----------------------- | -------------------------------------------------------------------------------------------- |
| `id`                    | Unique slug; must start with a lowercase letter, then lowercase letters, digits, `-`, or `_` |
| `title`                 | Human-readable name                                                                          |
| `description`           | Agent purpose (shown in the tool description exposed to the parent)                          |
| `llm.base_uri`          | OpenAI-compatible API base URL                                                               |
| `llm.api_key`           | API key; supports `${ENV_VAR}` substitution                                                  |
| `llm.model_id`          | Model identifier for the provider                                                            |
| `llm.reasoning_effort`  | Optional reasoning budget: `none`, `minimal`, `low`, `medium`, `high`, `xhigh`               |
| `llm.reasoning_summary` | Optional reasoning summary style: `auto`, `concise`, `detailed` (use `detailed` for longer)  |
| `llm.verbosity`         | Optional response verbosity for reasoning models: `low`, `medium`, `high`                      |
| `llm.max_tokens`        | Optional max output tokens (increase if answers are cut short)                               |
| `llm.temperature`       | Optional sampling temperature (`0.0`–`2.0`)                                                  |
| `system_prompt`         | System message for the sub-agent                                                             |
| `mcp_servers`           | List of remote MCP servers (`transport` must be `streamable_http`)                           |
| `mcp_servers[].name`    | Short name used in qualified tool names (`name.tool`)                                        |
| `mcp_servers[].url`     | Streamable HTTP MCP endpoint (must end with `/mcp` for standard layouts)                     |
| `mcp_servers[].bearer_token` | Optional bearer token; sent as `Authorization: Bearer ...` (supports `${ENV_VAR}`)      |
| `mcp_servers[].headers` | Optional extra HTTP headers merged with bearer auth                                          |
| `tool_allowlist`        | Optional list of `server.tool` names; omit to allow all tools from connected servers         |

Environment variable substitution supports `${VAR}` and `${VAR:-default}`. If `VAR` is unset and no default is provided, startup fails with a clear error.

### Configuration cookbook

**Second agent (different role, no extra MCP servers)**

```yaml
- id: writer
  title: Writing Agent
  description: "Drafts and edits text"
  llm:
    base_uri: https://api.openai.com/v1
    api_key: ${OPENAI_API_KEY}
    model_id: gpt-4.1-mini
  system_prompt: |
    You are a concise technical writer.
  mcp_servers: []
```

**Local model via Ollama (OpenAI-compatible)**

```yaml
llm:
  base_uri: http://localhost:11434/v1
  api_key: ${OLLAMA_API_KEY:-ollama}
  model_id: llama3.2
```

**Authenticated downstream MCP server**

```yaml
mcp_servers:
  - name: my_api
    transport: streamable_http
    url: https://mcp.example.com/mcp
    bearer_token: ${MY_MCP_TOKEN}
```

You can also set auth manually via `headers`:

```yaml
    headers:
      Authorization: "Bearer ${MY_MCP_TOKEN}"
```

**Allow all tools from connected servers** — Remove `tool_allowlist` or set it to `null` in YAML (omit the key).

### Startup validation errors

| Error                        | Typical cause                                                         |
| ---------------------------- | --------------------------------------------------------------------- |
| Config file not found        | Missing `config/agents.yaml`; copy from `agents.example.yaml`         |
| Environment variable not set | `${VAR}` without value or `${VAR:-}` default                          |
| Pydantic validation failed   | Invalid `id`, duplicate ids, empty `system_prompt`, wrong `transport` |
| Duplicate agent ids          | Two agents share the same `id`                                        |

## MCP tools reference

Each agent in `config/agents.yaml` is exposed as an MCP tool named by its `id`.

### Per-agent tools (for example, `researcher`)

| Parameter | Type   | Description                    |
| --------- | ------ | ------------------------------ |
| `prompt`  | string | User message for the sub-agent |

**Returns**

| Shape                   | Meaning                                                             |
| ----------------------- | ------------------------------------------------------------------- |
| `{ "response": "..." }` | Success; final assistant text                                       |
| `{ "error": "..." }`    | Failure (LLM error, MCP connection error, and so on)                |

Tool descriptions include `title`, `description`, and `model_id` but never API keys.

Errors are returned in the result object; they do not crash the MCP server process.

## Environment variables

| Variable                | Default              | Description                                                                                   |
| ----------------------- | -------------------- | --------------------------------------------------------------------------------------------- |
| `AGENTS_CONFIG_PATH`    | `config/agents.yaml` | Path to agents YAML                                                                           |
| `HOST`                  | `0.0.0.0`            | Server bind host                                                                              |
| `PORT`                  | `8000`               | Server bind port                                                                              |
| `LOG_LEVEL`             | `INFO`               | Log level (`DEBUG`, `INFO`, …)                                                                |
| `MCP_CLIENT_TIMEOUT`    | `30`                 | Timeout in seconds when connecting to downstream MCP servers; increase for slow tools         |
| `AGENT_RECURSION_LIMIT` | `25`                 | Maximum LangChain agent tool-loop steps per agent tool call; increase for multi-step tasks |

## Project layout

```
config/
  agents.example.yaml    # Template; copy to agents.yaml
  agents.yaml            # Runtime config (example committed in repo)
src/sub_agent_mcp/
  main.py                # FastMCP entry point
  server/                # Per-agent MCP tools, OpenAPI route
  agent/                 # LangChain builder and executor
  config/                # YAML loader and Pydantic schema
  mcp_client/            # Downstream MCP connections and tool registry
docker/mock_mcp/         # Dev mock filesystem and search MCP servers
tests/                   # pytest suite
scripts/bump-tag.sh      # Release tag helper (v0.major.minor)
```

## Development

```bash
uv sync --dev
uv run pytest -v
uv run ruff check src tests
```

Pull requests and pushes to `main` run lint and tests in [`.github/workflows/ci.yml`](.github/workflows/ci.yml).

## Docker image and releases

**Registry:** `ghcr.io/systemgroupnet/sub-agent-mcp`

**When images publish:** Pushing a git tag `v0.*` (for example `v0.1.2`) runs the Docker job after tests pass. Pushes to `main` alone do not publish images.

**Zero versioning:** Tags use `v0.major.minor` (for example `v0.1.0`, `v0.1.1`, `v0.2.0`).

In VS Code: **Terminal → Run Task → Release: bump zero-version tag and push**

Or manually:

```bash
./scripts/bump-tag.sh minor   # v0.1.0 → v0.1.1
./scripts/bump-tag.sh major   # v0.1.0 → v0.2.0
```

Image tags include the semver, `latest`, and major.minor aliases per the metadata action in CI.

## Troubleshooting

| Symptom                                  | Likely cause                                 | Fix                                                                                       |
| ---------------------------------------- | -------------------------------------------- | ----------------------------------------------------------------------------------------- |
| `Configuration error` on startup         | Missing or invalid `agents.yaml`             | Copy [config/agents.example.yaml](config/agents.example.yaml); check `AGENTS_CONFIG_PATH` |
| `Environment variable 'X' is not set`    | `${X}` without default in YAML               | `export X=...` or use `${X:-default}`                                                     |
| Agent tool MCP / connection errors       | Wrong MCP URL for your environment           | Use the [Local vs Docker MCP URLs](#local-vs-docker-mcp-urls) table                       |
| `401` / invalid credentials              | Bad `llm.api_key` for that agent             | Verify provider key and `base_uri`                                                        |
| Downstream MCP tools unavailable         | Mock servers not running or strict allowlist | Start `filesystem-mcp` / `search-mcp`; review `tool_allowlist`                            |
| Cursor cannot connect                    | Server not running or port blocked           | Confirm `curl http://localhost:8000/mcp/openapi.json`; check firewall / WSL networking    |
| Agent stops after few tool calls         | Recursion limit                              | Raise `AGENT_RECURSION_LIMIT`                                                             |

## License

[MIT](LICENSE)

Sub-Agent-MCP

Sub-Agent-MCP Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Sub-Agent-MCP

Sub-Agent-MCP Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README