Open Source

Building in
Public

The Equilibrium Coach ecosystem: open-source tools for AI-powered Dota 2 coaching and analysis.

Explore

Our Projects

Production-grade open-source libraries published on PyPI.

v1.4.7
01

python-manta

272 callbacks · CGO bindings

High-performance Dota 2 replay parser built with CGO bindings. Parses .dem files with 272 game event callbacks, extracting player actions, combat logs, and match metadata.

PythonGoCGODota 2
v7.40.3
02

python-opendota-sdk

Full API coverage · Async

Fully async OpenDota API wrapper with complete endpoint coverage. Pydantic models for type-safe responses, automatic rate limiting, and pagination support.

PythonAsyncPydanticAPI
v1.2.0
03

mcp-replay-dota2

FastMCP · AI Analysis

FastMCP server that enables AI models to analyze Dota 2 replays through the Model Context Protocol. Provides structured match data for coaching insights.

PythonMCPFastMCPAI
dev
04

dota2-coach-plugin

20+ MCP tools · Claude Code

Claude Code plugin that turns your terminal into a Dota 2 coaching assistant. 20+ MCP tools for replay analysis, hero matchup data, and personalized improvement plans.

TypeScriptClaude CodeMCPPlugin
Philosophy

Why We Build in Public

Open source is how we prove our engineering quality. Every library in this ecosystem was built using the same spec-driven workflow we use for client projects. The code speaks for itself.

Explore Our Work

Check out our GitHub organization or book a call to discuss how we can build for you.