Apache Maka: a high-performance agent workspace that keeps a complete record of everything it did (5.8k stars)
What it does
All text below is quoted directly from the project README.
>Apache Maka (Incubating) is a high-performance agent workspace that keeps a complete record of everything it did.
What Maka is:
- Measured, not claimed. Maka is benchmarked against other harnesses on the same model with the official verifier, and the per-task results ship with every report in docs/eval/.
- The log is the runtime. Every model message, tool call, permission decision and termination is an append-only RuntimeEvent. The UI, the next prompt and crash recovery are projections of that log, never the only copy. Old tool output can leave the next prompt without leaving the log.
- Your machine, your model. Sessions, settings and run records stay local. You bring the model: a cloud API, a local model or a compatible gateway.
- One Runtime Host. Desktop, the TUI and CLI, and Eval are thin clients of one execution authority; Eval owns only the experiment and its scores.
>npm run dev starts the Desktop development environment with HMR.
Repository
apache/maka
Repository stats
|Metric|Value|
|:-|:-|
|Stars|5,811|
|Forks|552|
|Last release|None|
|Languages|TypeScript 92.1%, JavaScript 4.8%, Rust 1.4%|
|Contributors|151|
|AI contributor accounts (Claude / Codex / Cursor / Copilot)|Copilot: 2, codex: 4, cursoragent: 5, claude: 79|
|AI co-authored commits|Claude: 79, Codex: 4, Cursor: 5, Copilot: 3|
Main contributor
|Field|Value|
|:-|:-|
|Name|jakevin|
|GitHub|@jackwener|
|Commits to this repo|1,550 (top contributor)|
|Bio|Building OpenCLI & Maka & Raft. AI enthusiast! Love Open Source! PMC member & Committer of @apache Arrow & Datafusion & Doris|
|Company / Location|@apache / Hangzhou, China|
|X / Twitter|@jakevin7|
|Followers|3123|
|Public repos|48|
|On GitHub since|2017-07-28|
31 points
Mentions
Maka 从 8 月开始也提出 Workhub 的概念，也就是由一个前端的 inbox bot 对所有需求进行分拣，交给不同的 Session 执行。 包括 Claude 的 Project，还有 Cursor 也有类似的设计，大家不约而同的走向了去 Session 的这个方向。 maka#3286
@kepano @obsdmd Apache Maka just switched Markdown diagrams from Mermaid.js to Pele 🌋 - Much faster rendering- Smaller JS bundle- Dropped ~100 dependency packages from Mermaid’s tree Theme changes no longer force re-renders either. Big win for us. apache/maka t.co
Apache Maka: a high-performance agent workspace that keeps a complete record of everything it did (5.8k stars) What it does All text below is quoted directly from the project README. >Apache Maka (Incubating) is a high-performance agent workspace that keeps a complete record of everything it did. What Maka is: - Measured, not claimed. Maka is benchmarked against other harnesses on the same model with the official verifier, and the per-task results ship with every report in docs/eval/. - The log is the runtime. Every model message, tool call, permission decision and termination is an append-only RuntimeEvent. The UI, the next prompt and crash recovery are projections of that log, never the only copy. Old tool output can leave the next prompt without leaving the log. - Your machine, your model. Sessions, settings and run records stay local. You bring the model: a cloud API, a local model or a compatible gateway. - One Runtime Host. Desktop, the TUI and CLI, and Eval are thin clients of one execution authority; Eval owns only the experiment and its scores. >npm run dev starts the Desktop development environment with HMR. Repository apache/maka Repository stats |Metric|Value| |:-|:-| |Stars|5,811| |Forks|552| |Last release|None| |Languages|TypeScript 92.1%, JavaScript 4.8%, Rust 1.4%| |Contributors|151| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|Copilot: 2, codex: 4, cursoragent: 5, claude: 79| |AI co-authored commits|Claude: 79, Codex: 4, Cursor: 5, Copilot: 3| Main contributor |Field|Value| |:-|:-| |Name|jakevin| |GitHub|@jackwener| |Commits to this repo|1,550 (top contributor)| |Bio|Building OpenCLI & Maka & Raft. AI enthusiast! Love Open Source! PMC member & Committer of @apache Arrow & Datafusion & Doris| |Company / Location|@apache / Hangzhou, China| |X / Twitter|@jakevin7| |Followers|3123| |Public repos|48| |On GitHub since|2017-07-28|
我很好奇这种大规模重写，Agent 真的能做得好决策吗？ 我最近放手纯 auto drive 了几个项目之后，发现即便是 alint-dev/alint 做规则约束也压不住代码腐化和过早决策带来的架构过拟合 稍微真的遇到一点问题亲自瞄一眼代码之后会发现哪哪都是错的
现在几乎主要的 Agent 都是用 Electron 方案，Codex、Claude...一旦同时跑的 Session 和工作多了，性能消耗和资源占用都很大，尤其是 Web 渲染的各种动画。 我们在把整个 Maka 重写为 Rust + GPUI。为未来几十/上百/上千个 agent 并行做好准备。 已经几乎可用了： github.com/apache/maka/tree/rust-preview
先日、この「記録を残す」ことを主眼にしたエージェントハーネス、 "Apache Maka" ってやつを見つけた。5.7kスター。 ほんの少し触っただけだけど、ChatGPTのデスクトップアプリみたいな感じ。トークン効率も重視されているらしい。プロバイダーも色々対応してました。 apache/maka