Skip to main content
Glama
KooshaPari

dispatch-mcp

by KooshaPari
OverviewSchemaRelated ServersScoreDiscussions
Python
Remote

Downloads GitHub release License AI-Slop AI-Only-Maintained HITL-less

⚠️ AI-Agent-Only Repository

This repo is planned, maintained, and managed exclusively by AI Agents. Slop issues, rough edges, and AI artifacts are expected and intentionally present as part of an HITL-less / minimized AI-DD metaproject focused on learning, refining, and brute-force training both the agents and the human operator. Bug reports and contributions are still welcome, but please expect AI-generated code, comments, and documentation throughout.

Work State

Field

Value

Last commit

2026-06-08 16:14:41 -0700

Open issues

0

Open PRs

3

Focus

omniroute-dispatch

Progress: ████████░░ 80%

dispatch-mcp

MCP server for tier-based dispatch delegation via OmniRoute.

Related MCP server: SystemPrompt MCP TaskChecker

Tools

Per-tier dispatch tools

Tool name

Tier

dispatch_worker

worker

dispatch_main

main

dispatch_codeman

codeman

dispatch_freetier

freetier

dispatch_kimi

kimi

dispatch_kimi_thinking

kimi_thinking

dispatch_minimax

minimax

dispatch_opus

opus

dispatch_haiku

haiku

dispatch_gemini

gemini

Each accepts a single message: str argument and dispatches it to the configured OmniRoute backend under the corresponding tier.

Custom dispatch

dispatch_custom(tier: str, message: str) — dispatch to any tier from VALID_TIERS above.

Health

  • dispatch_health() — probe the OmniRoute backend health endpoint. Requires OMNIROUTE_URL to be set.

  • dispatch_liveness() — returns server liveness status without contacting OmniRoute.

Health module (production)

dispatch_mcp.health exposes three pure-Python functions intended for production probes and observability:

Function

Purpose

I/O

liveness()

Cheap process-liveness probe (kubelet livenessProbe).

None

readiness(*, check_omniroute=False, timeout=2.0)

Readiness probe. Validates OMNIROUTE_URL config; opt-in upstream check via check_omniroute=True.

Optional HTTP GET /health

metrics()

Prometheus text exposition placeholder with dispatch_mcp_up, dispatch_mcp_uptime_seconds, and dispatch counter gauges.

None

Wire them into FastMCP tools, an HTTP wrapper, or invoke them directly from tests. Default readiness() is config-only to avoid stampeding the backend.

Configuration

Variable

Required

Default

Description

OMNIROUTE_URL

Yes

—

Base URL of the OmniRoute dispatch backend (e.g. http://localhost:8080). Must use http:// or https:// scheme.

LOG_LEVEL

No

(root logger)

Logging verbosity. Accepted values: DEBUG, INFO, WARNING, ERROR, CRITICAL. Invalid values fall through to the root logger's level.

Constraints

  • message must not exceed 4096 bytes (UTF-8 encoded).

  • tier must be one of the known tiers listed above.

  • OMNIROUTE_URL must use http:// or https:// scheme. Other schemes (e.g. file://, javascript:) are rejected at startup with a ValueError.

  • HTTP redirects are not followed — only direct requests to OMNIROUTE_URL are made.

Build

Install the package in editable mode with the dev extras (pulls in pytest, pytest-cov):

python -m pip install -e ".[dev]"

The runtime dependencies are fastmcp>=3.2.4 and httpx>=0.27.0. Requires Python 3.13+.

Test

# Run the full test suite with coverage (uses pyproject.toml addopts)
pytest

# Or with an explicit report
pytest --cov-report=term-missing

The default addopts (see [tool.pytest.ini_options]) enables branch coverage, missing-line reporting, and fails the run if coverage drops below 80%.

Lint & type check

# Format
ruff format .

# Lint
ruff check .

# Strict type check
mypy src/ --strict

The same checks run in CI (see .github/workflows/ci.yml).

Run

# Set the backend URL
export OMNIROUTE_URL=http://localhost:20128

# Via entry point
dispatch-mcp

# Or directly
python -m dispatch_mcp.server

This server cannot be installed

F
license - not found
Not graded
quality - not tested
B
maintenance

How are these scores calculated?

Maintenance

–Maintainers
–Response time
–Release cycle
–Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • Straight Connect

    AI-Guru
    F
    license
    Not graded
    quality
    D
    maintenance
    An MCP server for communication service connectors that currently provides multi-account Telegram integration with granular tool access and security controls. It allows AI models to manage messages, chats, and media across various accounts through a flexible, extensible routing architecture.
    1

  • Multi-LLM Gateway MCP

    tgosoul2019
    A
    license
    Not graded
    quality
    C
    maintenance
    An MCP server that functions as an intelligent gateway for multiple LLM backends including OpenAI, Claude, and Ollama. It supports automatic provider fallback, streaming responses via Server-Sent Events, and real-time monitoring for robust AI integration.
    MIT

View all related MCP servers

Related MCP Connectors

  • omnidim-mcp-server

    Official MCP server for OmniDimension. Drive voice agents, dispatch calls, and run bulk campaigns.

  • mcp

    MCP server connecting AI agents to non-custodial staking data across 130+ networks.

  • microservices.sh

    MCP server for AI agents to plan, verify, and deploy Cloudflare-native apps.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/KooshaPari/dispatch-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server