cytostack/

openwolf

GitHub Website

OpenWolf acts as a context management layer for AI coding agents, reusing a shared brain across multiple agents to optimize token usage. It provides a local dashboard, token accounting, and lifecycle hooks to manage context state.

2.2kstars

190forks

38issues

AGPL-3.0license

2026since

Star historydaily snapshots by VibeCrowd

Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).

Alternatives & relatedmatched by topic overlap

rtk-ai/

rtk

▲ 998/wk

CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies

75k4.7kRust

wshobson/

agents

▲ 148/wk

Multi-harness agentic plugin marketplace for Claude Code, Codex CLI, Cursor, OpenCode, GitHub Copilot, and Gemini CLI

38k4.1kPython

code-yeongyu/

oh-my-openagent

▲ 483/wk

omo/lazycodex: The coding agent for tokenmaxxers;the one and only agent harness for complex codebases. For your Codex, for your OpenCode

67k5.5kTypeScript

DeusData/

codebase-memory-mcp

▲ 1.2k/wk

High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

37k3.0kC

affaan-m/

ECC

▲ 2.9k/wk

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

238k36kJavaScript

OpenHands/

OpenHands

▲ 681/wk

🙌 OpenHands: AI-Driven Development

83k11kTypeScript

Compareopenwolf vs rtk openwolf vs agents openwolf vs oh-my-openagent

Reviewgenerated from repository data · Aug 5, 2026

What it is

OpenWolf is described as "The second brain for Claude Code. Now for every AI coding assistant." It focuses on improved context management, architecture scaffolding, and token utilization, delivered through 7 invisible lifecycle hooks.

How it works

The README explains a workflow where OpenWolf injects a token-budgeted digest at session start, then coordinates reads and symbol-level hints to minimize token usage. It maintains a durable project index with descriptions, token estimates, and content hashes, and uses a PreCompact hook to preserve state across context compaction. It logs actions and estimates costs, updating the memory and token ledger.

Getting started

npm install -g openwolf
cd your-project
openwolf init

This wires agents to a shared .wolf/ brain. Alternative invocations show wiring specific agents or all detected agents:

openwolf init                          # auto-detect installed agents (recommended)
openwolf init --agent codex opencode   # wire exactly these
openwolf init --agent all              # wire every detected agent
openwolf init --agent claude           # Claude Code only

Recent releases

v2.0.1 OpenWolf 2.0.1 (2026-07-14): Bug-fix patch. Fixed the dashboard white screen when a token is rejected; now shows a clear "token rejected" message. Root cause: the status bad
v2.0.0 OpenWolf 2.0.0 (2026-07-14): # OpenWolf 2.0.0 The second brain for Claude Code is now a context layer for every AI coding assistant. OpenWolf 2.0 emphasizes improved context management, opti

Traction

2171 stars, 190 forks, 38 open issues as of the data provided.

Behind the repo

The repository is authored under AGPL-3.0 and targets Node.js 20+, TypeScript 5.7. It supports 6 agents and provides multiple protocol blocks for integrating with Codex CLI, OpenCode, Claude Code, Cursor, Antigravity, and Gemini CLI.

Caveats

License: AGPL-3.0
Created: 2026-03-15; last push: 2026-07-15
Language: TypeScript
Requirements: Node.js 20+, at least one supported coding agent

SharePost on X LinkedIn