dgf1988/codewhale
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

docs(brand): rename to codewhale across READMEs and docs

Browse Source 
Sweep brand mentions of `DeepSeek TUI` / `deepseek-tui` / bare
`deepseek` (the dispatcher binary) across all user-facing docs to
the new `codewhale` brand. The DeepSeek **provider** integration is
left untouched throughout: env vars (`DEEPSEEK_*`), model IDs
(`deepseek-v4-pro`, `deepseek-v4-flash`, `deepseek-chat`,
`deepseek-reasoner`), the `api.deepseek.com` host, the
`~/.deepseek/` config dir, and the `--provider deepseek` argument
value all keep the legacy spelling.

Anti-scope items deliberately left as the legacy `deepseek-tui`:

- Homebrew tap and formula (`Hmbown/homebrew-deepseek-tui`,
  `brew install deepseek-tui`, `scoop install deepseek-tui`). The
  tap rename ships separately.
- Docker image (`ghcr.io/hmbown/deepseek-tui`). Image-tag rename
  ships separately.
- CNB mirror namespace (`cnb.cool/deepseek-tui.com/DeepSeek-TUI`).
  Third-party hosted path.
- Security contact email (`security@deepseek-tui.com`).
- GitHub repo URL (`Hmbown/DeepSeek-TUI`).

New artifact:

- `docs/REBRAND.md` documents what changed, what didn't, the
  deprecation window, and migration commands for npm / Cargo /
  Homebrew / manual installs.

CHANGELOG entries:

- Root `CHANGELOG.md` and `crates/tui/CHANGELOG.md` both gain a
  new `[Unreleased]` section describing the rename and the one-
  release deprecation window. Historical entries are untouched.

Issue templates:

- `.github/ISSUE_TEMPLATE/bug_report.md` and `feature_request.md`
  refer to "codewhale" / `codewhale --version` instead of the old
  brand name in their environment fields.

The rebrand sweep was driven by a perl script with bulk patterns
(`deepseek-tui` -> `codewhale-tui`, `DeepSeek TUI` -> `codewhale`,
bare `deepseek` -> `codewhale` with provider/model/host/env-var/
config-path negative lookbehind/lookahead) followed by targeted
reverts for the anti-scope items above. Output was visually
reviewed file-by-file before committing.

Verified:

- `cargo check --workspace --all-targets --locked` — pass.
- `cargo test --workspace --all-features --locked` — pass (no
  test source touched here; suite stayed green to confirm no
  doc-from-string assertions broke).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Hunter Bown
2026-05-23 11:25:48 -05:00
parent 23daefbe24
commit a3acdbe70b
25 changed files with 760 additions and 571 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/ISSUE_TEMPLATE/bug_report.md
+2 -2
View File
@@ -27,7 +27,7 @@ labels: bug
## Environment
- OS:
- DeepSeek CLI version:
- codewhale version:
- Install method:
- Model/provider:
- Terminal app:
@@ -35,7 +35,7 @@ labels: bug
<!-- Hints:
OS: Windows 11 / Ubuntu 22.04 / macOS 14
DeepSeek CLI version: run `deepseek --version`
codewhale version: run `codewhale --version`
Install method: cargo install / release binary / source build
Model/provider: deepseek-v4-pro / DeepSeek, or qwen2.5-coder / Ollama
Terminal app: iTerm2 / Windows Terminal / GNOME Terminal / VS Code terminal
.github/ISSUE_TEMPLATE/feature_request.md
+1 -1
View File
@@ -10,7 +10,7 @@ labels: enhancement
## Proposed solution
<!-- What should DeepSeek TUI do differently? Include commands, UI behavior, or config shape if relevant. -->
<!-- What should codewhale do differently? Include commands, UI behavior, or config shape if relevant. -->
## Use case
CHANGELOG.md
+25
View File
@@ -5,6 +5,31 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased]
### Changed
- **Project renamed to codewhale.** The canonical CLI dispatcher is now
`codewhale` (was `deepseek`) and the TUI runtime is `codewhale-tui`
(was `deepseek-tui`). The 14 workspace crates are renamed from
`deepseek-*` / `deepseek-tui-*` to `codewhale-*` / `codewhale-tui-*`.
The npm wrapper package is now `codewhale` (was `deepseek-tui`). See
[docs/REBRAND.md](docs/REBRAND.md) for migration notes.
- **DeepSeek provider integration is unchanged.** `DEEPSEEK_*` env vars,
model IDs (`deepseek-v4-pro`, `deepseek-v4-flash`, the legacy
`deepseek-chat` / `deepseek-reasoner` aliases), the
`https://api.deepseek.com` host, and the `~/.deepseek/` config
directory are all preserved.
### Deprecated
- The `deepseek` and `deepseek-tui` binary names continue to ship as
tiny shims that print a one-line warning and forward argv to the
renamed binaries. They will be removed in v0.9.0.
- The `deepseek-tui` npm package continues to publish for one release
cycle as a no-`bin` deprecation shim whose postinstall directs users
to `npm install -g codewhale`. It will be removed in v0.9.0.
## [0.8.40] - 2026-05-21
### Added
CONTRIBUTING.md
+7 -7
View File
@@ -1,6 +1,6 @@
# Contributing to DeepSeek TUI
# Contributing to codewhale
Thank you for your interest in contributing to DeepSeek TUI! This document provides guidelines and instructions for contributing.
Thank you for your interest in contributing to codewhale! This document provides guidelines and instructions for contributing.
## Getting Started
@@ -118,14 +118,14 @@ instead of the Harvest path, the highest-leverage things you can do are:
## Project Structure
DeepSeek TUI is a Cargo workspace. The live runtime and the majority of TUI,
codewhale is a Cargo workspace. The live runtime and the majority of TUI,
engine, and tool code currently live in `crates/tui/src/`. Smaller workspace
crates provide shared abstractions that are being extracted incrementally.
```
crates/
├── tui/ deepseek-tui binary (interactive TUI + runtime API)
├── cli/ deepseek binary (dispatcher facade)
├── tui/ codewhale-tui binary (interactive TUI + runtime API)
├── cli/ codewhale binary (dispatcher facade)
├── app-server/ HTTP/SSE + JSON-RPC transport
├── core/ Agent loop / session / turn management
├── protocol/ Request/response framing
@@ -201,7 +201,7 @@ When reporting issues, please include:
- Operating system and version
- Rust version (`rustc --version`)
- DeepSeek TUI version (`deepseek --version`)
- codewhale version (`codewhale --version`)
- Steps to reproduce the issue
- Expected vs actual behavior
- Relevant error messages or logs
@@ -212,7 +212,7 @@ Be respectful and inclusive. We welcome contributors of all backgrounds and expe
## License
By contributing to DeepSeek TUI, you agree that your contributions will be licensed under the MIT License.
By contributing to codewhale, you agree that your contributions will be licensed under the MIT License.
## Questions?
README.ja-JP.md
+76 -76
View File
@@ -1,4 +1,4 @@
# 🐳 DeepSeek TUI
# 🐳 codewhale
> **このターミナルネイティブのコーディングエージェントは、DeepSeek V4 の 100 万トークンのコンテキストウィンドウとプレフィックスキャッシュ機能を中心に構築されています。単一のバイナリとして配布され、Node.js や Python のランタイムは不要です。MCP クライアント、サンドボックス、永続的なタスクキューも標準で同梱されています。**
@@ -7,17 +7,17 @@
## インストール
`deepseek` は自己完結型の Rust バイナリとして提供されており、**実行に Node.js や Python のランタイムは必要ありません。** すでにマシンにインストールされているものを選んでください。いずれの方法でも同じバイナリが `PATH` に配置されます。
`codewhale` は自己完結型の Rust バイナリとして提供されており、**実行に Node.js や Python のランタイムは必要ありません。** すでにマシンにインストールされているものを選んでください。いずれの方法でも同じバイナリが `PATH` に配置されます。
```bash
# 1. npm — すでに Node を使っているなら最も簡単。npm パッケージは
# GitHub Releases から対応するビルド済みバイナリをダウンロードする
# 薄いインストーラーであり、deepseek 本体に Node ランタイム依存を加えるものではありません。
npm install -g deepseek-tui
# 薄いインストーラーであり、codewhale 本体に Node ランタイム依存を加えるものではありません。
npm install -g codewhale
# 2. Cargo — Node 不要。
cargo install deepseek-tui-cli --locked # `deepseek` (エントリーポイント)
cargo install deepseek-tui --locked # `deepseek-tui` (TUI バイナリ)
cargo install codewhale-cli --locked # `codewhale` (エントリーポイント)
cargo install codewhale-tui --locked # `codewhale-tui` (TUI バイナリ)
# 3. Homebrew — macOS パッケージマネージャ。
brew tap Hmbown/deepseek-tui
@@ -28,10 +28,10 @@ brew install deepseek-tui
# Linux x64/ARM64、macOS x64/ARM64、Windows x64 向けのビルド済みバイナリがあります。
# 5. Docker — ビルド済みリリースイメージ。
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -42,27 +42,27 @@ docker run --rm -it \
既にインストール済みの場合は、インストール方法に合わせて更新してください:
```bash
deepseek update
npm install -g deepseek-tui@latest
codewhale update
npm install -g codewhale@latest
brew update && brew upgrade deepseek-tui
cargo install deepseek-tui-cli --locked --force
cargo install deepseek-tui --locked --force
cargo install codewhale-cli --locked --force
cargo install codewhale-tui --locked --force
```
[![CI](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml/badge.svg)](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml)
[![npm](https://img.shields.io/npm/v/deepseek-tui)](https://www.npmjs.com/package/deepseek-tui)
[![crates.io](https://img.shields.io/crates/v/deepseek-tui-cli?label=crates.io)](https://crates.io/crates/deepseek-tui-cli)
[![npm](https://img.shields.io/npm/v/codewhale)](https://www.npmjs.com/package/codewhale)
[![crates.io](https://img.shields.io/crates/v/codewhale-cli?label=crates.io)](https://crates.io/crates/codewhale-cli)
[![DeepWiki](https://img.shields.io/badge/DeepWiki-Ask_AI-_.svg?style=flat&color=0052D9&labelColor=000000&logo=data:image/png;base64,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)](https://deepwiki.com/Hmbown/DeepSeek-TUI)
<a href="https://www.buymeacoffee.com/hmbown" target="_blank"><img src="https://img.shields.io/badge/Buy%20me%20a%20coffee-5F7FFF?style=for-the-badge&logo=buymeacoffee&logoColor=white" alt="Buy me a coffee" /></a>
![DeepSeek TUI スクリーンショット](assets/screenshot.png)
![codewhale スクリーンショット](assets/screenshot.png)
---
## DeepSeek TUI とは？
## codewhale とは？
DeepSeek TUI は、ターミナル内で完結するコーディングエージェントです。DeepSeek のフロンティアモデルがあなたのワークスペースに直接アクセスできるようにし、ファイルの読み取り・編集、シェルコマンドの実行、Web 検索、Git 管理、サブエージェントの統制などを、すべて高速でキーボード駆動の TUI を通じて行えます。
codewhale は、ターミナル内で完結するコーディングエージェントです。DeepSeek のフロンティアモデルがあなたのワークスペースに直接アクセスできるようにし、ファイルの読み取り・編集、シェルコマンドの実行、Web 検索、Git 管理、サブエージェントの統制などを、すべて高速でキーボード駆動の TUI を通じて行えます。
**DeepSeek V4 向けに構築** (`deepseek-v4-pro` / `deepseek-v4-flash`)。100 万トークンのコンテキストウィンドウとネイティブの thinking-mode（思考連鎖）ストリーミングをサポートします。
@@ -78,7 +78,7 @@ DeepSeek TUI は、ターミナル内で完結するコーディングエージ
- **セッション保存／再開** — 長時間実行のセッションをチェックポイント化して再開可能
- **ワークスペースのロールバック** — リポジトリの `.git` には触れずに、サイド Git によるターン前後のスナップショットを `/restore` と `revert_turn` で扱える
- **永続的タスクキュー** — 再起動を超えて生き残るバックグラウンドタスク。スケジュール自動化や長時間レビューなどに
- **HTTP/SSE ランタイム API** — `deepseek serve --http` でヘッドレスエージェントワークフローを実現
- **HTTP/SSE ランタイム API** — `codewhale serve --http` でヘッドレスエージェントワークフローを実現
- **MCP プロトコル** — Model Context Protocol サーバーに接続して拡張ツールを利用可能。詳細は [docs/MCP.md](docs/MCP.md) を参照
- **LSP 診断** — rust-analyzer、pyright、typescript-language-server、gopls、clangd により、編集ごとにエラー／警告をインライン表示
- **ユーザーメモリ** — クロスセッションの嗜好をシステムプロンプトに注入できる、オプションの永続メモファイル
@@ -90,7 +90,7 @@ DeepSeek TUI は、ターミナル内で完結するコーディングエージ
## 仕組み
`deepseek`（ディスパッチャー CLI）→ `deepseek-tui`（コンパニオンバイナリ）→ ratatui インターフェース ↔ 非同期エンジン ↔ OpenAI 互換のストリーミングクライアント。ツール呼び出しは型付きレジストリ（シェル、ファイル操作、Git、Web、サブエージェント、MCP、RLM）を経由してルーティングされ、結果はトランスクリプトへとストリーム返送されます。エンジンはセッション状態、ターン管理、永続タスクキューを管理し、LSP サブシステムは編集後の診断を次の推論ステップ前にモデルのコンテキストへ供給します。
`codewhale`（ディスパッチャー CLI）→ `codewhale-tui`（コンパニオンバイナリ）→ ratatui インターフェース ↔ 非同期エンジン ↔ OpenAI 互換のストリーミングクライアント。ツール呼び出しは型付きレジストリ（シェル、ファイル操作、Git、Web、サブエージェント、MCP、RLM）を経由してルーティングされ、結果はトランスクリプトへとストリーム返送されます。エンジンはセッション状態、ターン管理、永続タスクキューを管理し、LSP サブシステムは編集後の診断を次の推論ステップ前にモデルのコンテキストへ供給します。
詳しくは [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md) を参照してください。
@@ -99,9 +99,9 @@ DeepSeek TUI は、ターミナル内で完結するコーディングエージ
## クイックスタート
```bash
npm install -g deepseek-tui
deepseek --version
deepseek --model auto
npm install -g codewhale
codewhale --version
codewhale --model auto
```
ビルド済みバイナリは **Linux x64**、**Linux ARM64**（v0.8.8 以降）、**macOS x64**、**macOS ARM64**、**Windows x64** 向けに公開されています。その他のターゲット（musl、riscv64、FreeBSD など）は [ソースからのインストール](#install-from-source) または [docs/INSTALL.md](docs/INSTALL.md) を参照してください。
@@ -111,19 +111,19 @@ deepseek --model auto
事前に設定することもできます:
```bash
deepseek auth set --provider deepseek # ~/.deepseek/config.toml に保存
codewhale auth set --provider deepseek # ~/.deepseek/config.toml に保存
export DEEPSEEK_API_KEY="YOUR_KEY" # 環境変数による代替方法。非対話シェルでは ~/.zshenv を使用
deepseek
codewhale
deepseek doctor # セットアップを検証
codewhale doctor # セットアップを検証
```
> 保存済みキーをローテーション／削除するには: `deepseek auth clear --provider deepseek`。
> 保存済みキーをローテーション／削除するには: `codewhale auth clear --provider deepseek`。
### Linux ARM64（Raspberry Pi、Asahi、Graviton、HarmonyOS PC）
`npm i -g deepseek-tui` は v0.8.8 以降、glibc ベースの ARM64 Linux で動作します。[Releases ページ](https://github.com/Hmbown/DeepSeek-TUI/releases) からビルド済みバイナリをダウンロードし、`PATH` 上に並べて配置することもできます。
`npm i -g codewhale` は v0.8.8 以降、glibc ベースの ARM64 Linux で動作します。[Releases ページ](https://github.com/Hmbown/DeepSeek-TUI/releases) からビルド済みバイナリをダウンロードし、`PATH` 上に並べて配置することもできます。
### 中国 / ミラーフレンドリーなインストール
@@ -141,9 +141,9 @@ registry = "sparse+https://mirrors.tuna.tsinghua.edu.cn/crates.io-index/"
その後、両方のバイナリをインストールしてください（ディスパッチャーは実行時に TUI へ委譲します）:
```bash
cargo install deepseek-tui-cli --locked # `deepseek` を提供
cargo install deepseek-tui --locked # `deepseek-tui` を提供
deepseek --version
cargo install codewhale-cli --locked # `codewhale` を提供
cargo install codewhale-tui --locked # `codewhale-tui` を提供
codewhale --version
```
ビルド済みバイナリは [GitHub Releases](https://github.com/Hmbown/DeepSeek-TUI/releases) からもダウンロードできます。ミラーされたリリースアセットには `DEEPSEEK_TUI_RELEASE_BASE_URL` を使ってください。
@@ -170,8 +170,8 @@ scoop install deepseek-tui
git clone https://github.com/Hmbown/DeepSeek-TUI.git
cd DeepSeek-TUI
cargo install --path crates/cli --locked # Rust 1.88+ が必要。`deepseek` を提供
cargo install --path crates/tui --locked # `deepseek-tui` を提供
cargo install --path crates/cli --locked # Rust 1.88+ が必要。`codewhale` を提供
cargo install --path crates/tui --locked # `codewhale-tui` を提供
```
両方のバイナリが必要です。クロスコンパイルとプラットフォーム固有の注意事項: [docs/INSTALL.md](docs/INSTALL.md)。
@@ -182,42 +182,42 @@ cargo install --path crates/tui --locked # `deepseek-tui` を提供
```bash
# NVIDIA NIM
deepseek auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
deepseek --provider nvidia-nim
codewhale auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
codewhale --provider nvidia-nim
# AtlasCloud
deepseek auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
deepseek --provider atlascloud
codewhale auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
codewhale --provider atlascloud
# Wanjie Ark
deepseek auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
deepseek --provider wanjie-ark --model deepseek-reasoner
codewhale auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
codewhale --provider wanjie-ark --model deepseek-reasoner
# OpenRouter
deepseek auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
deepseek --provider openrouter --model deepseek/deepseek-v4-pro
codewhale auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
codewhale --provider openrouter --model deepseek/deepseek-v4-pro
# Novita
deepseek auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
deepseek --provider novita --model deepseek/deepseek-v4-pro
codewhale auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
codewhale --provider novita --model deepseek/deepseek-v4-pro
# Fireworks
deepseek auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
deepseek --provider fireworks --model deepseek-v4-pro
codewhale auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
codewhale --provider fireworks --model deepseek-v4-pro
# 汎用 OpenAI 互換エンドポイント
deepseek auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" deepseek --provider openai --model glm-5
codewhale auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" codewhale --provider openai --model glm-5
# セルフホスト SGLang
SGLANG_BASE_URL="http://localhost:30000/v1" deepseek --provider sglang --model deepseek-v4-flash
SGLANG_BASE_URL="http://localhost:30000/v1" codewhale --provider sglang --model deepseek-v4-flash
# セルフホスト vLLM
VLLM_BASE_URL="http://localhost:8000/v1" deepseek --provider vllm --model deepseek-v4-flash
VLLM_BASE_URL="http://localhost:8000/v1" codewhale --provider vllm --model deepseek-v4-flash
# セルフホスト Ollama
ollama pull deepseek-coder:1.3b
deepseek --provider ollama --model deepseek-coder:1.3b
ollama pull codewhale-coder:1.3b
codewhale --provider ollama --model codewhale-coder:1.3b
```
TUI 内では `/provider` でプロバイダーピッカー、`/model` でローカルのモデル/思考モードピッカーを開けます。`/provider openrouter` や `/model <id>` で直接切り替え、`/models` で対応プロバイダーのライブモデル一覧を明示的に取得できます。
@@ -233,30 +233,30 @@ TUI 内では `/provider` でプロバイダーピッカー、`/model` でロー
## 使い方
```bash
deepseek # インタラクティブ TUI
deepseek "explain this function" # ワンショットプロンプト
deepseek exec --auto --output-format stream-json "fix this bug" # NDJSON バックエンドストリーム
deepseek exec --resume <SESSION_ID> "follow up" # 非対話セッションを継続
deepseek --model deepseek-v4-flash "summarize" # モデルの上書き
deepseek --model auto "fix this bug" # モデルと推論強度を自動選択
deepseek --yolo # ツールを自動承認
deepseek auth set --provider deepseek # API キーの保存
deepseek doctor # セットアップと接続性のチェック
deepseek doctor --json # 機械可読の診断
deepseek setup --status # 読み取り専用のセットアップ状態
deepseek setup --tools --plugins # ツール／プラグインディレクトリの雛形作成
deepseek models # ライブ API モデル一覧
deepseek sessions # 保存済みセッション一覧
deepseek resume --last # 最新セッションを再開
deepseek resume <SESSION_ID> # UUID 指定で特定セッションを再開
deepseek fork <SESSION_ID> # 保存済みセッションを兄弟パスに fork
deepseek serve --http # HTTP/SSE API サーバー
deepseek serve --acp # Zed/カスタムエージェント向け ACP stdio アダプター
deepseek run pr <N> # PR を取得しレビュープロンプトに先行投入
deepseek mcp list # 設定された MCP サーバー一覧
deepseek mcp validate # MCP の設定／接続性を検証
deepseek mcp-server # ディスパッチャー MCP stdio サーバーを実行
deepseek update # バイナリ更新の確認と適用
codewhale # インタラクティブ TUI
codewhale "explain this function" # ワンショットプロンプト
codewhale exec --auto --output-format stream-json "fix this bug" # NDJSON バックエンドストリーム
codewhale exec --resume <SESSION_ID> "follow up" # 非対話セッションを継続
codewhale --model deepseek-v4-flash "summarize" # モデルの上書き
codewhale --model auto "fix this bug" # モデルと推論強度を自動選択
codewhale --yolo # ツールを自動承認
codewhale auth set --provider deepseek # API キーの保存
codewhale doctor # セットアップと接続性のチェック
codewhale doctor --json # 機械可読の診断
codewhale setup --status # 読み取り専用のセットアップ状態
codewhale setup --tools --plugins # ツール／プラグインディレクトリの雛形作成
codewhale models # ライブ API モデル一覧
codewhale sessions # 保存済みセッション一覧
codewhale resume --last # 最新セッションを再開
codewhale resume <SESSION_ID> # UUID 指定で特定セッションを再開
codewhale fork <SESSION_ID> # 保存済みセッションを兄弟パスに fork
codewhale serve --http # HTTP/SSE API サーバー
codewhale serve --acp # Zed/カスタムエージェント向け ACP stdio アダプター
codewhale run pr <N> # PR を取得しレビュープロンプトに先行投入
codewhale mcp list # 設定された MCP サーバー一覧
codewhale mcp validate # MCP の設定／接続性を検証
codewhale mcp-server # ディスパッチャー MCP stdio サーバーを実行
codewhale update # バイナリ更新の確認と適用
```
### キーボードショートカット
@@ -302,7 +302,7 @@ deepseek update # バイナリ更新の確認
| `DEEPSEEK_HTTP_HEADERS` | 任意のモデルリクエストヘッダー |
| `DEEPSEEK_MODEL` | デフォルトモデル |
| `DEEPSEEK_STREAM_IDLE_TIMEOUT_SECS` | ストリームのアイドルタイムアウト秒数 |
| `DEEPSEEK_PROVIDER` | `deepseek`（デフォルト）、`nvidia-nim`、`openai`、`atlascloud`、`wanjie-ark`、`openrouter`、`novita`、`fireworks`、`sglang`、`vllm`、`ollama` |
| `DEEPSEEK_PROVIDER` | `codewhale`（デフォルト）、`nvidia-nim`、`openai`、`atlascloud`、`wanjie-ark`、`openrouter`、`novita`、`fireworks`、`sglang`、`vllm`、`ollama` |
| `DEEPSEEK_PROFILE` | 設定プロファイル名 |
| `DEEPSEEK_MEMORY` | `on` に設定するとユーザーメモリを有効化 |
| `DEEPSEEK_ALLOW_INSECURE_HTTP=1` | 信頼できるネットワークで非ローカル `http://` API ベース URL を許可 |
@@ -341,7 +341,7 @@ UI のロケールはモデルの言語とは別です。`settings.toml` で `lo
## 自分のスキルを公開する
DeepSeek TUI はワークスペースのディレクトリ（`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills`）とグローバルな `~/.deepseek/skills` からスキルを発見します。各スキルは `SKILL.md` ファイルを持つディレクトリです:
codewhale はワークスペースのディレクトリ（`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills`）とグローバルな `~/.deepseek/skills` からスキルを発見します。各スキルは `SKILL.md` ファイルを持つディレクトリです:
```text
~/.deepseek/skills/my-skill/
README.md
+95 -95
View File
@@ -1,14 +1,14 @@
# DeepSeek TUI
# codewhale
> Terminal coding agent for DeepSeek V4. It runs from the `deepseek` command, streams reasoning blocks, edits local workspaces with approval gates, and includes an auto mode that chooses both model and thinking level per turn.
> Terminal coding agent for DeepSeek V4. It runs from the `codewhale` command, streams reasoning blocks, edits local workspaces with approval gates, and includes an auto mode that chooses both model and thinking level per turn.
[简体中文 README](README.zh-CN.md)
[日本語 README](README.ja-JP.md)
## Install
`deepseek` is distributed as Rust binaries: the dispatcher command
(`deepseek`) and the companion TUI runtime (`deepseek-tui`). Pick whichever
`codewhale` is distributed as Rust binaries: the dispatcher command
(`codewhale`) and the companion TUI runtime (`codewhale-tui`). Pick whichever
install path you already use; they all put the same commands on your `PATH`.
The npm package is an installer/wrapper for the release binaries, not the
agent runtime itself.
@@ -16,13 +16,13 @@ agent runtime itself.
```bash
# 1. npm — easiest if you already use Node. The package downloads the
# matching prebuilt Rust binaries from GitHub Releases.
npm install -g deepseek-tui
npm install -g codewhale
# 2. Cargo — no Node needed. Requires Rust 1.88+ (the crates use the
# 2024 edition; older toolchains fail with "feature `edition2024` is
# required"). Run `rustup update` first, or use a non-Cargo path below.
cargo install deepseek-tui-cli --locked # `deepseek` (entry point)
cargo install deepseek-tui --locked # `deepseek-tui` (TUI binary)
cargo install codewhale-cli --locked # `codewhale` (entry point)
cargo install codewhale-tui --locked # `codewhale-tui` (TUI binary)
# 3. Homebrew — macOS package manager.
brew tap Hmbown/deepseek-tui
@@ -33,10 +33,10 @@ brew install deepseek-tui
# Prebuilt for Linux x64/ARM64, macOS x64/ARM64, Windows x64.
# 5. Docker — prebuilt release image.
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -54,25 +54,25 @@ docker run --rm -it \
Already installed? Use the updater that matches the install path:
```bash
deepseek update # release-binary updater
npm install -g deepseek-tui@latest # npm wrapper
codewhale update # release-binary updater
npm install -g codewhale@latest # npm wrapper
brew update && brew upgrade deepseek-tui
cargo install deepseek-tui-cli --locked --force
cargo install deepseek-tui --locked --force
cargo install codewhale-cli --locked --force
cargo install codewhale-tui --locked --force
```
[![CI](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml/badge.svg)](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml)
[![npm](https://img.shields.io/npm/v/deepseek-tui)](https://www.npmjs.com/package/deepseek-tui)
[![crates.io](https://img.shields.io/crates/v/deepseek-tui-cli?label=crates.io)](https://crates.io/crates/deepseek-tui-cli)
[![npm](https://img.shields.io/npm/v/codewhale)](https://www.npmjs.com/package/codewhale)
[![crates.io](https://img.shields.io/crates/v/codewhale-cli?label=crates.io)](https://crates.io/crates/codewhale-cli)
[DeepWiki project index](https://deepwiki.com/Hmbown/DeepSeek-TUI)
![DeepSeek TUI screenshot](assets/screenshot.png)
![codewhale screenshot](assets/screenshot.png)
---
## What Is It?
DeepSeek TUI is a coding agent that runs in your terminal. It can read and edit files, run shell commands, search the web, manage git, and coordinate sub-agents from a keyboard-driven TUI.
codewhale is a coding agent that runs in your terminal. It can read and edit files, run shell commands, search the web, manage git, and coordinate sub-agents from a keyboard-driven TUI.
It is built around DeepSeek V4 (`deepseek-v4-pro` / `deepseek-v4-flash`), including 1M-token context windows, streaming reasoning blocks, and prefix-cache-aware cost reporting.
@@ -89,7 +89,7 @@ It is built around DeepSeek V4 (`deepseek-v4-pro` / `deepseek-v4-flash`), includ
- **Workspace rollback** — side-git pre/post-turn snapshots with `/restore` and `revert_turn`, without touching your repo's `.git`
- **OS-level sandbox** — Seatbelt on macOS, Landlock on Linux, Job Objects on Windows; shell commands run with workspace-scoped filesystem access only
- **Durable task queue** — background tasks can survive restarts
- **HTTP/SSE runtime API** — `deepseek serve --http` for headless agent workflows
- **HTTP/SSE runtime API** — `codewhale serve --http` for headless agent workflows
- **MCP protocol** — connect to Model Context Protocol servers for extended tooling; please see [docs/MCP.md](docs/MCP.md)
- **Native RLM** (`rlm_open`/`rlm_eval`) — persistent REPL sessions for batched analysis; run cheap `deepseek-v4-flash` children with bounded helpers like `peek`, `search`, `chunk`, and `sub_query_batch`
- **LSP diagnostics** — inline error/warning surfacing after every edit via rust-analyzer, pyright, typescript-language-server, gopls, clangd
@@ -104,17 +104,17 @@ It is built around DeepSeek V4 (`deepseek-v4-pro` / `deepseek-v4-flash`), includ
## How It's Wired
`deepseek` (dispatcher CLI) → `deepseek-tui` (companion binary) → ratatui interface ↔ async engine ↔ OpenAI-compatible streaming client. Tool calls route through a typed registry (shell, file ops, git, web, sub-agents, MCP, RLM) and results stream back into the transcript. The engine manages session state, turn tracking, the durable task queue, and an LSP subsystem that feeds post-edit diagnostics into the model's context before the next reasoning step.
`codewhale` (dispatcher CLI) → `codewhale-tui` (companion binary) → ratatui interface ↔ async engine ↔ OpenAI-compatible streaming client. Tool calls route through a typed registry (shell, file ops, git, web, sub-agents, MCP, RLM) and results stream back into the transcript. The engine manages session state, turn tracking, the durable task queue, and an LSP subsystem that feeds post-edit diagnostics into the model's context before the next reasoning step.
See [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md) for the full walkthrough.
### Sub-agents: Concurrent Background Execution
DeepSeek TUI can dispatch multiple sub-agents that run in parallel — like a concurrent task queue:
codewhale can dispatch multiple sub-agents that run in parallel — like a concurrent task queue:
- **Non-blocking launch.** `agent_open` returns immediately. The child gets its own fresh context and tool registry and runs independently. The parent keeps working.
- **Background execution.** Sub-agents execute concurrently (default cap: 10, configurable to 20). The engine manages the pool — no polling loop needed.
- **Completion notification.** When a sub-agent finishes, the runtime delivers a structured `<deepseek:subagent.done>` event with a summary, evidence list, and execution metrics. The parent model reads the `summary` field and integrates findings.
- **Completion notification.** When a sub-agent finishes, the runtime delivers a structured `<codewhale:subagent.done>` event with a summary, evidence list, and execution metrics. The parent model reads the `summary` field and integrates findings.
- **Bounded result retrieval.** Large transcripts are parked behind `var_handle` references. The model calls `handle_read` for slices, ranges, or JSONPath projections — keeping the parent context lean.
See [docs/SUBAGENTS.md](docs/SUBAGENTS.md) for the full sub-agent reference.
@@ -124,9 +124,9 @@ See [docs/SUBAGENTS.md](docs/SUBAGENTS.md) for the full sub-agent reference.
## Quickstart
```bash
npm install -g deepseek-tui
deepseek --version
deepseek --model auto
npm install -g codewhale
codewhale --version
codewhale --model auto
```
Prebuilt binaries are published for **Linux x64**, **Linux ARM64** (v0.8.8+), **macOS x64**, **macOS ARM64**, and **Windows x64**. For other targets (musl, riscv64, FreeBSD, etc.), see [Install from source](#install-from-source) or [docs/INSTALL.md](docs/INSTALL.md).
@@ -136,22 +136,22 @@ On first launch you'll be prompted for your [DeepSeek API key](https://platform.
You can also set it ahead of time:
```bash
deepseek auth set --provider deepseek # saves to ~/.deepseek/config.toml
deepseek auth status # shows the active credential source
codewhale auth set --provider deepseek # saves to ~/.deepseek/config.toml
codewhale auth status # shows the active credential source
export DEEPSEEK_API_KEY="YOUR_KEY" # env var alternative; use ~/.zshenv for non-interactive shells
deepseek
codewhale
deepseek doctor # verify setup
codewhale doctor # verify setup
```
If `deepseek doctor` says the rejected key came from `DEEPSEEK_API_KEY`, remove
If `codewhale doctor` says the rejected key came from `DEEPSEEK_API_KEY`, remove
the stale export from your shell startup file, open a fresh shell, or run
`deepseek auth set --provider deepseek`. Use `deepseek auth status` to see the
`codewhale auth set --provider deepseek`. Use `codewhale auth status` to see the
config, keyring, and env-var source state without printing the key. Saved config
keys take precedence over the keyring and environment and are easier to rotate.
> To rotate or remove a saved key: `deepseek auth clear --provider deepseek`.
> To rotate or remove a saved key: `codewhale auth clear --provider deepseek`.
### Tencent Cloud / CNB Remote-First Path
@@ -166,7 +166,7 @@ server runbook.
### Auto Mode
Use `deepseek --model auto` or `/model auto` when you want DeepSeek TUI to decide how much model and reasoning power a turn needs.
Use `codewhale --model auto` or `/model auto` when you want codewhale to decide how much model and reasoning power a turn needs.
Auto mode controls two settings together:
@@ -175,13 +175,13 @@ Auto mode controls two settings together:
Before the real turn is sent, the app makes a small `deepseek-v4-flash` routing call with thinking off. That router looks at the latest request and recent context, then selects a concrete model and thinking level for the real request. Short/simple turns can stay on Flash with thinking off; coding, debugging, release work, architecture, security review, or ambiguous multi-step tasks can move up to Pro and/or higher thinking.
`auto` is local to DeepSeek TUI. The upstream API never receives `model: "auto"`; it receives the concrete model and thinking setting chosen for that turn. The TUI shows the selected route, and cost tracking is charged against the model that actually ran. If the router call fails or returns an invalid answer, the app falls back to a local heuristic. Sub-agents inherit auto mode unless you assign them an explicit model.
`auto` is local to codewhale. The upstream API never receives `model: "auto"`; it receives the concrete model and thinking setting chosen for that turn. The TUI shows the selected route, and cost tracking is charged against the model that actually ran. If the router call fails or returns an invalid answer, the app falls back to a local heuristic. Sub-agents inherit auto mode unless you assign them an explicit model.
Use a fixed model or fixed thinking level when you want repeatable benchmarking, a strict cost ceiling, or a specific provider/model mapping.
### Linux ARM64 (Raspberry Pi, Asahi, Graviton, HarmonyOS PC)
`npm i -g deepseek-tui` works on glibc-based ARM64 Linux from v0.8.8 onward. You can also download prebuilt binaries from the [Releases page](https://github.com/Hmbown/DeepSeek-TUI/releases) and place them side by side on your `PATH`.
`npm i -g codewhale` works on glibc-based ARM64 Linux from v0.8.8 onward. You can also download prebuilt binaries from the [Releases page](https://github.com/Hmbown/DeepSeek-TUI/releases) and place them side by side on your `PATH`.
### China / Mirror-friendly Installation
@@ -199,24 +199,24 @@ registry = "sparse+https://mirrors.tuna.tsinghua.edu.cn/crates.io-index/"
Then install both binaries (the dispatcher delegates to the TUI at runtime):
```bash
cargo install deepseek-tui-cli --locked # provides `deepseek`
cargo install deepseek-tui --locked # provides `deepseek-tui`
deepseek --version
cargo install codewhale-cli --locked # provides `codewhale`
cargo install codewhale-tui --locked # provides `codewhale-tui`
codewhale --version
```
Prebuilt binaries can also be downloaded from [GitHub Releases](https://github.com/Hmbown/DeepSeek-TUI/releases). Use `DEEPSEEK_TUI_RELEASE_BASE_URL` for mirrored release assets.
### Windows (Scoop)
[Scoop](https://scoop.sh) is a Windows package manager. DeepSeek TUI is listed
[Scoop](https://scoop.sh) is a Windows package manager. codewhale is listed
in Scoop's main bucket, but that manifest updates independently and can lag the
GitHub/npm/Cargo release. Run `scoop update` first, then verify the installed
version with `deepseek --version`:
version with `codewhale --version`:
```bash
scoop update
scoop install deepseek-tui
deepseek --version
codewhale --version
```
Use npm or direct GitHub release downloads when you need the newest release
@@ -236,8 +236,8 @@ Works on any Tier-1 Rust target — including musl, riscv64, FreeBSD, and older
git clone https://github.com/Hmbown/DeepSeek-TUI.git
cd DeepSeek-TUI
cargo install --path crates/cli --locked # requires Rust 1.88+; provides `deepseek`
cargo install --path crates/tui --locked # provides `deepseek-tui`
cargo install --path crates/cli --locked # requires Rust 1.88+; provides `codewhale`
cargo install --path crates/tui --locked # provides `codewhale-tui`
```
Both binaries are required. Cross-compilation and platform-specific notes: [docs/INSTALL.md](docs/INSTALL.md).
@@ -248,42 +248,42 @@ Both binaries are required. Cross-compilation and platform-specific notes: [docs
```bash
# NVIDIA NIM
deepseek auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
deepseek --provider nvidia-nim
codewhale auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
codewhale --provider nvidia-nim
# AtlasCloud
deepseek auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
deepseek --provider atlascloud
codewhale auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
codewhale --provider atlascloud
# Wanjie Ark
deepseek auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
deepseek --provider wanjie-ark --model deepseek-reasoner
codewhale auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
codewhale --provider wanjie-ark --model deepseek-reasoner
# OpenRouter
deepseek auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
deepseek --provider openrouter --model deepseek/deepseek-v4-pro
codewhale auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
codewhale --provider openrouter --model deepseek/deepseek-v4-pro
# Novita
deepseek auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
deepseek --provider novita --model deepseek/deepseek-v4-pro
codewhale auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
codewhale --provider novita --model deepseek/deepseek-v4-pro
# Fireworks
deepseek auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
deepseek --provider fireworks --model deepseek-v4-pro
codewhale auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
codewhale --provider fireworks --model deepseek-v4-pro
# Generic OpenAI-compatible endpoint
deepseek auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" deepseek --provider openai --model glm-5
codewhale auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" codewhale --provider openai --model glm-5
# Self-hosted SGLang
SGLANG_BASE_URL="http://localhost:30000/v1" deepseek --provider sglang --model deepseek-v4-flash
SGLANG_BASE_URL="http://localhost:30000/v1" codewhale --provider sglang --model deepseek-v4-flash
# Self-hosted vLLM
VLLM_BASE_URL="http://localhost:8000/v1" deepseek --provider vllm --model deepseek-v4-flash
VLLM_BASE_URL="http://localhost:8000/v1" codewhale --provider vllm --model deepseek-v4-flash
# Self-hosted Ollama
ollama pull deepseek-coder:1.3b
deepseek --provider ollama --model deepseek-coder:1.3b
ollama pull codewhale-coder:1.3b
codewhale --provider ollama --model codewhale-coder:1.3b
```
Inside the TUI, `/provider` opens the provider picker and `/model` opens the
@@ -304,38 +304,38 @@ interfaces, and extension points.
## Usage
```bash
deepseek # interactive TUI
deepseek "explain this function" # one-shot prompt
deepseek exec --auto --output-format stream-json "fix this bug" # NDJSON backend stream
deepseek exec --resume <SESSION_ID> "follow up" # continue a non-interactive session
deepseek --model deepseek-v4-flash "summarize" # model override
deepseek --model auto "fix this bug" # auto-select model + thinking
deepseek --yolo # auto-approve tools
deepseek auth set --provider deepseek # save API key
deepseek doctor # check setup & connectivity
deepseek doctor --json # machine-readable diagnostics
deepseek setup --status # read-only setup status
deepseek setup --tools --plugins # scaffold tool/plugin dirs
deepseek models # list live API models
deepseek sessions # list saved sessions
deepseek resume --last # resume the most recent session in this workspace
deepseek resume <SESSION_ID> # resume a specific session by UUID
deepseek fork <SESSION_ID> # fork a saved session into a sibling path
deepseek serve --http # HTTP/SSE API server
deepseek serve --acp # ACP stdio adapter for Zed/custom agents
deepseek run pr <N> # fetch PR and pre-seed review prompt
deepseek mcp list # list configured MCP servers
deepseek mcp validate # validate MCP config/connectivity
deepseek mcp-server # run dispatcher MCP stdio server
deepseek update # check for and apply binary updates
codewhale # interactive TUI
codewhale "explain this function" # one-shot prompt
codewhale exec --auto --output-format stream-json "fix this bug" # NDJSON backend stream
codewhale exec --resume <SESSION_ID> "follow up" # continue a non-interactive session
codewhale --model deepseek-v4-flash "summarize" # model override
codewhale --model auto "fix this bug" # auto-select model + thinking
codewhale --yolo # auto-approve tools
codewhale auth set --provider deepseek # save API key
codewhale doctor # check setup & connectivity
codewhale doctor --json # machine-readable diagnostics
codewhale setup --status # read-only setup status
codewhale setup --tools --plugins # scaffold tool/plugin dirs
codewhale models # list live API models
codewhale sessions # list saved sessions
codewhale resume --last # resume the most recent session in this workspace
codewhale resume <SESSION_ID> # resume a specific session by UUID
codewhale fork <SESSION_ID> # fork a saved session into a sibling path
codewhale serve --http # HTTP/SSE API server
codewhale serve --acp # ACP stdio adapter for Zed/custom agents
codewhale run pr <N> # fetch PR and pre-seed review prompt
codewhale mcp list # list configured MCP servers
codewhale mcp validate # validate MCP config/connectivity
codewhale mcp-server # run dispatcher MCP stdio server
codewhale update # check for and apply binary updates
```
### Branching Conversations
Saved sessions are intentionally branchable. `deepseek fork <SESSION_ID>` copies
Saved sessions are intentionally branchable. `codewhale fork <SESSION_ID>` copies
an existing saved session into a new sibling session, records the parent session
id in metadata, and opens that fork so you can explore an alternate direction
without polluting the original path. The session picker and `deepseek sessions`
without polluting the original path. The session picker and `codewhale sessions`
mark forked sessions with their parent id.
Inside the TUI, Esc-Esc backtrack can rewind the active transcript to a prior
@@ -346,11 +346,11 @@ from side-git snapshots but do not rewrite conversation history.
Docker images are published to GHCR for release builds:
```bash
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -369,7 +369,7 @@ spawn local ACP agents over stdio. In Zed, add a custom agent server:
"agent_servers": {
"DeepSeek": {
"type": "custom",
"command": "deepseek",
"command": "codewhale",
"args": ["serve", "--acp"],
"env": {}
}
@@ -381,8 +381,8 @@ The first ACP slice supports new sessions and prompt responses through your
existing DeepSeek config/API key. Tool-backed editing and checkpoint replay are
not exposed through ACP yet.
Community-maintained adapter: [acp-deepseek-adapter](https://github.com/rockeverm3m/acp-deepseek-adapter)
bridges `deepseek exec --auto` to `cc-connect` for users who need tool-backed
Community-maintained adapter: [acp-codewhale-adapter](https://github.com/rockeverm3m/acp-codewhale-adapter)
bridges `codewhale exec --auto` to `cc-connect` for users who need tool-backed
ACP workflows outside the built-in Zed slice.
### Keyboard Shortcuts
@@ -427,7 +427,7 @@ Key environment variables:
| `DEEPSEEK_HTTP_HEADERS` | Optional custom model request headers, e.g. `X-Model-Provider-Id=your-model-provider` |
| `DEEPSEEK_MODEL` | Default model |
| `DEEPSEEK_STREAM_IDLE_TIMEOUT_SECS` | Stream idle timeout in seconds, default `300`, clamped to `1..=3600` |
| `DEEPSEEK_PROVIDER` | `deepseek` (default), `nvidia-nim`, `openai`, `atlascloud`, `wanjie-ark`, `openrouter`, `novita`, `fireworks`, `sglang`, `vllm`, `ollama` |
| `DEEPSEEK_PROVIDER` | `codewhale` (default), `nvidia-nim`, `openai`, `atlascloud`, `wanjie-ark`, `openrouter`, `novita`, `fireworks`, `sglang`, `vllm`, `ollama` |
| `DEEPSEEK_PROFILE` | Config profile name |
| `DEEPSEEK_MEMORY` | Set to `on` to enable user memory |
| `DEEPSEEK_ALLOW_INSECURE_HTTP=1` | Allow non-local `http://` API base URLs on trusted networks |
@@ -471,7 +471,7 @@ Legacy aliases `deepseek-chat` / `deepseek-reasoner` map to `deepseek-v4-flash`
## Publishing Your Own Skill
DeepSeek TUI discovers skills from workspace directories (`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills` → `.cursor/skills`) and global directories (`~/.agents/skills` → `~/.claude/skills` → `~/.deepseek/skills`). Each skill is a directory with a `SKILL.md` file:
codewhale discovers skills from workspace directories (`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills` → `.cursor/skills`) and global directories (`~/.agents/skills` → `~/.claude/skills` → `~/.deepseek/skills`). Each skill is a directory with a `SKILL.md` file:
```text
~/.agents/skills/my-skill/
@@ -530,7 +530,7 @@ Full Changelog: [CHANGELOG.md](CHANGELOG.md).
- **[DeepSeek](https://github.com/deepseek-ai)** — thank you for the models and support that power every turn. 感谢 DeepSeek 提供模型与支持，让每一次交互成为可能。
- **[DataWhale](https://github.com/datawhalechina)** 🐋 — thank you for your support and for welcoming us into the Whale Brother family. 感谢 DataWhale 的支持，并欢迎我们加入“鲸兄弟”大家庭。
- **[OpenWarp](https://github.com/zerx-lab/warp)** — thank you for prioritizing DeepSeek TUI support and for collaborating on a better terminal-agent experience.
- **[OpenWarp](https://github.com/zerx-lab/warp)** — thank you for prioritizing codewhale support and for collaborating on a better terminal-agent experience.
- **[Open Design](https://github.com/nexu-io/open-design)** — thank you for support and collaboration around design-forward agent workflows.
This project ships with help from a growing community of contributors:
@@ -556,7 +556,7 @@ This project ships with help from a growing community of contributors:
- **[xieshutao](https://github.com/xieshutao)** — plain Markdown skill fallback (#869)
- **[GK012](https://github.com/GK012)** — npm wrapper `--version` fallback (#885)
- **[y0sif](https://github.com/y0sif)** — parent turn-loop wakeup after direct child sub-agent completion (#901)
- **[mac119](https://github.com/mac119)** and **[leo119](https://github.com/leo119)** — `deepseek update` command documentation (#838, #917)
- **[mac119](https://github.com/mac119)** and **[leo119](https://github.com/leo119)** — `codewhale update` command documentation (#838, #917)
- **[dumbjack](https://github.com/dumbjack)** / **浩淼的mac** — command-safety null-byte hardening (#706, #918)
- **macworkers** — fork confirmation with the new session id (#600, #919)
- **zero** and **[zerx-lab](https://github.com/zerx-lab)** — notification condition config and richer OSC 9 notification body (#820, #920)
README.zh-CN.md
+89 -89
View File
@@ -1,4 +1,4 @@
# DeepSeek TUI
# codewhale
> **面向 [DeepSeek V4](https://platform.deepseek.com) 的终端原生编程智能体：100 万 token 上下文、思考模式流式推理、前缀缓存感知。自包含 Rust 二进制发布——开箱即带 MCP 客户端、沙箱和持久化任务队列。**
@@ -7,18 +7,18 @@
## 安装
`deepseek` 是自包含 Rust 二进制——**运行时不依赖 Node.js 或 Python**。
`codewhale` 是自包含 Rust 二进制——**运行时不依赖 Node.js 或 Python**。
下面几种方式装出来的是同一套二进制，按你已有的工具链选一个即可：
```bash
# 1. npm —— 已装 Node 的最方便方式。npm 包只是一个下载器，
# 会从 GitHub Releases 拉取对应平台的预编译二进制，
# 并不会让 deepseek 本身依赖 Node 运行时。
npm install -g deepseek-tui
# 并不会让 codewhale 本身依赖 Node 运行时。
npm install -g codewhale
# 2. Cargo —— 无需 Node。
cargo install deepseek-tui-cli --locked # `deepseek` 入口
cargo install deepseek-tui --locked # `deepseek-tui` TUI 二进制
cargo install codewhale-cli --locked # `codewhale` 入口
cargo install codewhale-tui --locked # `codewhale-tui` TUI 二进制
# 3. Homebrew —— macOS 包管理器。
brew tap Hmbown/deepseek-tui
@@ -29,10 +29,10 @@ brew install deepseek-tui
# 覆盖 Linux x64/ARM64、macOS x64/ARM64、Windows x64
# 5. Docker —— 预构建发布镜像。
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -49,25 +49,25 @@ docker run --rm -it \
已经安装过？按你的安装方式更新：
```bash
deepseek update # release 二进制更新器
npm install -g deepseek-tui@latest # npm 包装器
codewhale update # release 二进制更新器
npm install -g codewhale@latest # npm 包装器
brew update && brew upgrade deepseek-tui
cargo install deepseek-tui-cli --locked --force
cargo install deepseek-tui --locked --force
cargo install codewhale-cli --locked --force
cargo install codewhale-tui --locked --force
```
[![CI](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml/badge.svg)](https://github.com/Hmbown/DeepSeek-TUI/actions/workflows/ci.yml)
[![npm](https://img.shields.io/npm/v/deepseek-tui)](https://www.npmjs.com/package/deepseek-tui)
[![crates.io](https://img.shields.io/crates/v/deepseek-tui-cli?label=crates.io)](https://crates.io/crates/deepseek-tui-cli)
[![npm](https://img.shields.io/npm/v/codewhale)](https://www.npmjs.com/package/codewhale)
[![crates.io](https://img.shields.io/crates/v/codewhale-cli?label=crates.io)](https://crates.io/crates/codewhale-cli)
[DeepWiki project index](https://deepwiki.com/Hmbown/DeepSeek-TUI)
![DeepSeek TUI 截图](assets/screenshot.png)
![codewhale 截图](assets/screenshot.png)
---
## 这是什么？
DeepSeek TUI 是一个完全运行在终端里的编程智能体。它让 DeepSeek 前沿模型直接访问你的工作区：读写文件、运行 shell 命令、搜索浏览网页、管理 git、调度子智能体——全部通过快速、键盘驱动的 TUI 完成。
codewhale 是一个完全运行在终端里的编程智能体。它让 DeepSeek 前沿模型直接访问你的工作区：读写文件、运行 shell 命令、搜索浏览网页、管理 git、调度子智能体——全部通过快速、键盘驱动的 TUI 完成。
它面向 **DeepSeek V4**（`deepseek-v4-pro` / `deepseek-v4-flash`）构建，原生支持 100 万 token 上下文窗口和思考模式流式输出。
@@ -84,7 +84,7 @@ DeepSeek TUI 是一个完全运行在终端里的编程智能体。它让 DeepSe
- **会话保存和恢复** —— 长任务的断点续作
- **工作区回滚** —— 通过 side-git 记录每轮前后快照，支持 `/restore` 和 `revert_turn`，不影响项目自己的 `.git`
- **持久化任务队列** —— 后台任务在重启后仍然存在，支持计划任务和长时间运行的操作
- **HTTP/SSE 运行时 API** —— `deepseek serve --http` 用于无界面智能体流程
- **HTTP/SSE 运行时 API** —— `codewhale serve --http` 用于无界面智能体流程
- **MCP 协议** —— 连接 Model Context Protocol 服务器扩展工具，见 [docs/MCP.md](docs/MCP.md)
- **LSP 诊断** —— 每次编辑后通过 rust-analyzer、pyright、typescript-language-server、gopls、clangd 提供内联错误/警告
- **用户记忆** —— 可选的持久化笔记文件注入系统提示，实现跨会话偏好保持
@@ -98,17 +98,17 @@ DeepSeek TUI 是一个完全运行在终端里的编程智能体。它让 DeepSe
## 架构说明
`deepseek`（调度器 CLI）→ `deepseek-tui`（伴随二进制）→ ratatui 界面 ↔ 异步引擎 ↔ OpenAI 兼容流式客户端。工具调用通过类型化注册表（shell、文件操作、git、web、子智能体、MCP、RLM）路由，结果流式返回对话记录。引擎管理会话状态、轮次追踪、持久化任务队列和 LSP 子系统——它在下一步推理前将编辑后诊断反馈到模型上下文中。
`codewhale`（调度器 CLI）→ `codewhale-tui`（伴随二进制）→ ratatui 界面 ↔ 异步引擎 ↔ OpenAI 兼容流式客户端。工具调用通过类型化注册表（shell、文件操作、git、web、子智能体、MCP、RLM）路由，结果流式返回对话记录。引擎管理会话状态、轮次追踪、持久化任务队列和 LSP 子系统——它在下一步推理前将编辑后诊断反馈到模型上下文中。
详见 [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md)。
### 子智能体：并发后台执行
DeepSeek TUI 可以同时调度多个子智能体并行运行——类似于并发任务队列：
codewhale 可以同时调度多个子智能体并行运行——类似于并发任务队列：
- **非阻塞启动。** `agent_open` 立即返回。子智能体获得独立的上下文和工具注册表，独立运行。父进程继续工作。
- **后台执行。** 子智能体并发运行（默认上限 10，可配置至 20）。引擎管理线程池——无需轮询循环。
- **完成通知。** 子智能体完成后，运行时发送结构化的 `<deepseek:subagent.done>` 事件，包含摘要、证据列表和执行指标。父模型读取 `summary` 字段并整合结果。
- **完成通知。** 子智能体完成后，运行时发送结构化的 `<codewhale:subagent.done>` 事件，包含摘要、证据列表和执行指标。父模型读取 `summary` 字段并整合结果。
- **按需读取结果。** 大型对话记录暂存为 `var_handle` 引用。模型通过 `handle_read` 按切片、范围或 JSONPath 投影读取——保持父上下文精简。
详见 [docs/SUBAGENTS.md](docs/SUBAGENTS.md)。
@@ -118,9 +118,9 @@ DeepSeek TUI 可以同时调度多个子智能体并行运行——类似于并
## 快速开始
```bash
npm install -g deepseek-tui
deepseek --version
deepseek --model auto
npm install -g codewhale
codewhale --version
codewhale --model auto
```
预构建二进制覆盖 **Linux x64**、**Linux ARM64**（v0.8.8 起）、**macOS x64**、**macOS ARM64** 和 **Windows x64**。其他目标平台（musl、riscv64、FreeBSD 等）请见下方的[从源码安装](#从源码安装)或 [docs/INSTALL.md](docs/INSTALL.md)。
@@ -130,16 +130,16 @@ deepseek --model auto
也可以提前配置：
```bash
deepseek auth set --provider deepseek # 保存到 ~/.deepseek/config.toml
codewhale auth set --provider deepseek # 保存到 ~/.deepseek/config.toml
deepseek auth status # 显示当前活跃的凭证来源
codewhale auth status # 显示当前活跃的凭证来源
export DEEPSEEK_API_KEY="YOUR_KEY" # 环境变量方式；需要在非交互式 shell 中使用请放入 ~/.zshenv
deepseek
codewhale
deepseek doctor # 验证安装
codewhale doctor # 验证安装
```
> 轮换或移除密钥：`deepseek auth clear --provider deepseek`。
> 轮换或移除密钥：`codewhale auth clear --provider deepseek`。
### 腾讯云 / CNB 远程优先路径
@@ -153,7 +153,7 @@ CNB 镜像/源码，腾讯云 Lighthouse 香港实例，飞书/Lark 长连接桥
### Auto 模式
使用 `deepseek --model auto` 或 `/model auto` 让 DeepSeek TUI 自行决定每轮需要多少模型和推理能力。
使用 `codewhale --model auto` 或 `/model auto` 让 codewhale 自行决定每轮需要多少模型和推理能力。
Auto 模式同时控制两个设置：
@@ -162,13 +162,13 @@ Auto 模式同时控制两个设置：
在真实请求发出之前，应用会先用关闭推理的 `deepseek-v4-flash` 进行一次小型路由调用。路由器审视最新请求和最近的上下文，然后为真实请求选定具体的模型和推理强度。简短/简单的轮次保持在 Flash + 关闭推理；编码、调试、发布、架构、安全审查或模糊的多步骤任务可升级到 Pro 和/或更高推理强度。
`auto` 是 DeepSeek TUI 本地行为。上游 API 永远不会收到 `model: "auto"`，它只会收到为当前轮次选定的具体模型和推理强度设置。TUI 会显示选定的路由，成本跟踪按实际运行的模型计费。如果路由调用失败或返回无效答案，应用会回退到本地启发式规则。子智能体会继承 auto 模式，除非你为它们指定了显式模型。
`auto` 是 codewhale 本地行为。上游 API 永远不会收到 `model: "auto"`，它只会收到为当前轮次选定的具体模型和推理强度设置。TUI 会显示选定的路由，成本跟踪按实际运行的模型计费。如果路由调用失败或返回无效答案，应用会回退到本地启发式规则。子智能体会继承 auto 模式，除非你为它们指定了显式模型。
需要可重复基准测试、严格控制成本上限或特定提供商/模型映射时，请使用固定模型或固定推理强度。
### Linux ARM64（HarmonyOS 轻薄本、openEuler、Kylin、树莓派、Graviton 等）
从 v0.8.8 起，`npm i -g deepseek-tui` 直接支持 glibc 系的 ARM64 Linux。你也可以从 [Releases 页面](https://github.com/Hmbown/DeepSeek-TUI/releases) 下载预编译二进制，放到 `PATH` 目录中。
从 v0.8.8 起，`npm i -g codewhale` 直接支持 glibc 系的 ARM64 Linux。你也可以从 [Releases 页面](https://github.com/Hmbown/DeepSeek-TUI/releases) 下载预编译二进制，放到 `PATH` 目录中。
### 中国大陆 / 镜像友好安装
@@ -186,23 +186,23 @@ registry = "sparse+https://mirrors.tuna.tsinghua.edu.cn/crates.io-index/"
然后安装两个二进制（调度器在运行时会调用 TUI）：
```bash
cargo install deepseek-tui-cli --locked # 提供推荐入口 `deepseek`
cargo install deepseek-tui --locked # 提供交互式 TUI 伴随二进制
deepseek --version
cargo install codewhale-cli --locked # 提供推荐入口 `codewhale`
cargo install codewhale-tui --locked # 提供交互式 TUI 伴随二进制
codewhale --version
```
也可以直接从 [GitHub Releases](https://github.com/Hmbown/DeepSeek-TUI/releases) 下载预编译二进制。`DEEPSEEK_TUI_RELEASE_BASE_URL` 可用于镜像后的 release 资产。
### Windows (Scoop)
[Scoop](https://scoop.sh) 是一个 Windows 软件包管理器。DeepSeek TUI 已进入
[Scoop](https://scoop.sh) 是一个 Windows 软件包管理器。codewhale 已进入
Scoop main bucket，但该 manifest 独立更新，可能滞后于 GitHub/npm/Cargo
release。先运行 `scoop update`，安装后用 `deepseek --version` 核对版本：
release。先运行 `scoop update`，安装后用 `codewhale --version` 核对版本：
```bash
scoop update
scoop install deepseek-tui
deepseek --version
codewhale --version
```
如果需要最新版本，请优先使用 npm 或直接下载 GitHub Release 资产。
@@ -221,8 +221,8 @@ deepseek --version
git clone https://github.com/Hmbown/DeepSeek-TUI.git
cd DeepSeek-TUI
cargo install --path crates/cli --locked # 需要 Rust 1.88+；提供 `deepseek`
cargo install --path crates/tui --locked # 提供 `deepseek-tui`
cargo install --path crates/cli --locked # 需要 Rust 1.88+；提供 `codewhale`
cargo install --path crates/tui --locked # 提供 `codewhale-tui`
```
两个二进制都需要安装。交叉编译和平台特定说明见 [docs/INSTALL.md](docs/INSTALL.md)。
@@ -233,42 +233,42 @@ cargo install --path crates/tui --locked # 提供 `deepseek-tui`
```bash
# NVIDIA NIM
deepseek auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
deepseek --provider nvidia-nim
codewhale auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"
codewhale --provider nvidia-nim
# AtlasCloud
deepseek auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
deepseek --provider atlascloud
codewhale auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"
codewhale --provider atlascloud
# Wanjie Ark
deepseek auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
deepseek --provider wanjie-ark --model deepseek-reasoner
codewhale auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"
codewhale --provider wanjie-ark --model deepseek-reasoner
# OpenRouter
deepseek auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
deepseek --provider openrouter --model deepseek/deepseek-v4-pro
codewhale auth set --provider openrouter --api-key "YOUR_OPENROUTER_API_KEY"
codewhale --provider openrouter --model deepseek/deepseek-v4-pro
# Novita
deepseek auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
deepseek --provider novita --model deepseek/deepseek-v4-pro
codewhale auth set --provider novita --api-key "YOUR_NOVITA_API_KEY"
codewhale --provider novita --model deepseek/deepseek-v4-pro
# Fireworks
deepseek auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
deepseek --provider fireworks --model deepseek-v4-pro
codewhale auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"
codewhale --provider fireworks --model deepseek-v4-pro
# 通用 OpenAI 兼容端点
deepseek auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" deepseek --provider openai --model glm-5
codewhale auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"
OPENAI_BASE_URL="https://openai-compatible.example/v4" codewhale --provider openai --model glm-5
# 自托管 SGLang
SGLANG_BASE_URL="http://localhost:30000/v1" deepseek --provider sglang --model deepseek-v4-flash
SGLANG_BASE_URL="http://localhost:30000/v1" codewhale --provider sglang --model deepseek-v4-flash
# 自托管 vLLM
VLLM_BASE_URL="http://localhost:8000/v1" deepseek --provider vllm --model deepseek-v4-flash
VLLM_BASE_URL="http://localhost:8000/v1" codewhale --provider vllm --model deepseek-v4-flash
# 自托管 Ollama
ollama pull deepseek-coder:1.3b
deepseek --provider ollama --model deepseek-coder:1.3b
ollama pull codewhale-coder:1.3b
codewhale --provider ollama --model codewhale-coder:1.3b
```
在 TUI 内，`/provider` 打开提供方选择器，`/model` 打开本地模型/思考模式
@@ -287,40 +287,40 @@ deepseek --provider ollama --model deepseek-coder:1.3b
## 使用方式
```bash
deepseek # 交互式 TUI
deepseek "explain this function" # 一次性提示
deepseek exec --auto --output-format stream-json "fix this bug" # 面向后端集成的 NDJSON 流
deepseek exec --resume <SESSION_ID> "follow up" # 继续非交互会话
deepseek --model deepseek-v4-flash "summarize" # 指定模型
deepseek --model auto "fix this bug" # 自动选择模型 + 推理强度
deepseek --yolo # 自动批准工具
deepseek auth set --provider deepseek # 保存 API key
deepseek doctor # 检查配置和连接
deepseek doctor --json # 机器可读诊断
deepseek setup --status # 只读安装状态
deepseek setup --tools --plugins # 创建本地工具和插件目录
deepseek models # 列出可用 API 模型
deepseek sessions # 列出已保存会话
deepseek resume --last # 恢复最近会话
deepseek resume <SESSION_ID> # 按 UUID 恢复指定会话
deepseek fork <SESSION_ID> # 将已保存会话分叉为兄弟路径
deepseek serve --http # HTTP/SSE API 服务
deepseek serve --acp # Zed/自定义智能体的 ACP stdio 适配器
deepseek run pr <N> # 获取 PR 并预填审查提示
deepseek mcp list # 列出已配置 MCP 服务器
deepseek mcp validate # 校验 MCP 配置和连接
deepseek mcp-server # 启动 dispatcher MCP stdio 服务器
deepseek update # 检查并应用二进制更新
codewhale # 交互式 TUI
codewhale "explain this function" # 一次性提示
codewhale exec --auto --output-format stream-json "fix this bug" # 面向后端集成的 NDJSON 流
codewhale exec --resume <SESSION_ID> "follow up" # 继续非交互会话
codewhale --model deepseek-v4-flash "summarize" # 指定模型
codewhale --model auto "fix this bug" # 自动选择模型 + 推理强度
codewhale --yolo # 自动批准工具
codewhale auth set --provider deepseek # 保存 API key
codewhale doctor # 检查配置和连接
codewhale doctor --json # 机器可读诊断
codewhale setup --status # 只读安装状态
codewhale setup --tools --plugins # 创建本地工具和插件目录
codewhale models # 列出可用 API 模型
codewhale sessions # 列出已保存会话
codewhale resume --last # 恢复最近会话
codewhale resume <SESSION_ID> # 按 UUID 恢复指定会话
codewhale fork <SESSION_ID> # 将已保存会话分叉为兄弟路径
codewhale serve --http # HTTP/SSE API 服务
codewhale serve --acp # Zed/自定义智能体的 ACP stdio 适配器
codewhale run pr <N> # 获取 PR 并预填审查提示
codewhale mcp list # 列出已配置 MCP 服务器
codewhale mcp validate # 校验 MCP 配置和连接
codewhale mcp-server # 启动 dispatcher MCP stdio 服务器
codewhale update # 检查并应用二进制更新
```
Docker 镜像发布在 GHCR 上：
```bash
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -337,7 +337,7 @@ DeepSeek 可作为自定义 Agent Client Protocol 服务器运行，供 Zed 等
"agent_servers": {
"DeepSeek": {
"type": "custom",
"command": "deepseek",
"command": "codewhale",
"args": ["serve", "--acp"],
"env": {}
}
@@ -389,7 +389,7 @@ DeepSeek 可作为自定义 Agent Client Protocol 服务器运行，供 Zed 等
| `DEEPSEEK_HTTP_HEADERS` | 可选模型请求头，例如 `X-Model-Provider-Id=your-model-provider` |
| `DEEPSEEK_MODEL` | 默认模型 |
| `DEEPSEEK_STREAM_IDLE_TIMEOUT_SECS` | 流式响应空闲超时秒数，默认 `300`，限制在 `1..=3600` |
| `DEEPSEEK_PROVIDER` | `deepseek`（默认）、`nvidia-nim`、`openai`、`atlascloud`、`wanjie-ark`、`openrouter`、`novita`、`fireworks`、`sglang`、`vllm`、`ollama` |
| `DEEPSEEK_PROVIDER` | `codewhale`（默认）、`nvidia-nim`、`openai`、`atlascloud`、`wanjie-ark`、`openrouter`、`novita`、`fireworks`、`sglang`、`vllm`、`ollama` |
| `DEEPSEEK_PROFILE` | 配置 profile 名称 |
| `DEEPSEEK_MEMORY` | 设为 `on` 启用用户记忆 |
| `DEEPSEEK_ALLOW_INSECURE_HTTP=1` | 在可信网络上允许非本机 `http://` API base URL |
@@ -431,7 +431,7 @@ locale = "zh-Hans"
或者通过环境变量（中文系统通常已自动生效）：
```bash
LANG=zh_CN.UTF-8 deepseek run
LANG=zh_CN.UTF-8 codewhale run
```
---
@@ -454,7 +454,7 @@ LANG=zh_CN.UTF-8 deepseek run
## 创建和安装技能
DeepSeek TUI 从工作区目录（`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills`）和全局 `~/.deepseek/skills` 发现技能。每个技能是一个包含 `SKILL.md` 的目录：
codewhale 从工作区目录（`.agents/skills` → `skills` → `.opencode/skills` → `.claude/skills`）和全局 `~/.deepseek/skills` 发现技能。每个技能是一个包含 `SKILL.md` 的目录：
```text
~/.deepseek/skills/my-skill/
@@ -506,7 +506,7 @@ description: 当 DeepSeek 需要遵循我的自定义工作流时使用这个技
- **[DeepSeek](https://github.com/deepseek-ai)** — 感谢 DeepSeek 提供模型与支持，让每一次交互成为可能。
- **[DataWhale](https://github.com/datawhalechina)** — 感谢 DataWhale 的支持，并欢迎我们加入“鲸兄弟”大家庭。
- **[OpenWarp](https://github.com/zerx-lab/warp)** — 感谢 OpenWarp 优先支持 DeepSeek TUI，并一起打磨更好的终端智能体体验。
- **[OpenWarp](https://github.com/zerx-lab/warp)** — 感谢 OpenWarp 优先支持 codewhale，并一起打磨更好的终端智能体体验。
- **[Open Design](https://github.com/nexu-io/open-design)** — 感谢 Open Design 对面向设计的智能体工作流提供支持与协作。
本项目由不断壮大的贡献者社区共同打造：
@@ -532,7 +532,7 @@ description: 当 DeepSeek 需要遵循我的自定义工作流时使用这个技
- **[xieshutao](https://github.com/xieshutao)** — 纯 Markdown skill 兜底解析 (#869)
- **[GK012](https://github.com/GK012)** — npm wrapper 的 `--version` 兜底 (#885)
- **[y0sif](https://github.com/y0sif)** — 直接子智能体完成后唤醒父级 turn loop (#901)
- **[mac119](https://github.com/mac119)** 和 **[leo119](https://github.com/leo119)** — `deepseek update` 命令文档 (#838, #917)
- **[mac119](https://github.com/mac119)** 和 **[leo119](https://github.com/leo119)** — `codewhale update` 命令文档 (#838, #917)
- **[dumbjack](https://github.com/dumbjack)** / **浩淼的mac** — shell 命令空字节安全加固 (#706, #918)
- **macworkers** — fork 完成后显示新 session id (#600, #919)
- **zero** 和 **[zerx-lab](https://github.com/zerx-lab)** — 通知条件配置和更完整的 OSC 9 通知正文 (#820, #920)
SECURITY.md
+2 -2
View File
@@ -1,6 +1,6 @@
# Security Policy
DeepSeek TUI is a coding agent with direct access to file operations, shell execution, and the network. Security disclosures are taken seriously.
codewhale is a coding agent with direct access to file operations, shell execution, and the network. Security disclosures are taken seriously.
## Supported Versions
@@ -58,7 +58,7 @@ You will receive status updates at each phase. If the timeline slips, we will co
- Denial of service / rate-limit exhaustion against the DeepSeek API
- Vulnerabilities in third-party dependencies (report to the upstream project)
- Attacks requiring physical access to the victim's machine
- Theoretical ML-model injection attacks not demonstrated in the DeepSeek TUI context
- Theoretical ML-model injection attacks not demonstrated in the codewhale context
If you are unsure whether a bug is in scope, report it anyway. We will triage and respond.
crates/tui/CHANGELOG.md
+26
View File
@@ -5,6 +5,32 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased]
### Changed
- **Project renamed to codewhale.** The canonical CLI dispatcher is now
`codewhale` (was `deepseek`) and the TUI runtime is `codewhale-tui`
(was `deepseek-tui`). The 14 workspace crates are renamed from
`deepseek-*` / `deepseek-tui-*` to `codewhale-*` / `codewhale-tui-*`.
The npm wrapper package is now `codewhale` (was `deepseek-tui`). See
[docs/REBRAND.md](https://github.com/Hmbown/DeepSeek-TUI/blob/main/docs/REBRAND.md)
for migration notes.
- **DeepSeek provider integration is unchanged.** `DEEPSEEK_*` env vars,
model IDs (`deepseek-v4-pro`, `deepseek-v4-flash`, the legacy
`deepseek-chat` / `deepseek-reasoner` aliases), the
`https://api.deepseek.com` host, and the `~/.deepseek/` config
directory are all preserved.
### Deprecated
- The `deepseek` and `deepseek-tui` binary names continue to ship as
tiny shims that print a one-line warning and forward argv to the
renamed binaries. They will be removed in v0.9.0.
- The `deepseek-tui` npm package continues to publish for one release
cycle as a no-`bin` deprecation shim whose postinstall directs users
to `npm install -g codewhale`. It will be removed in v0.9.0.
## [0.8.40] - 2026-05-21
### Added
docs/ARCHITECTURE.md
+3 -3
View File
@@ -1,6 +1,6 @@
# DeepSeek TUI Architecture
# codewhale Architecture
This document provides an overview of the DeepSeek TUI architecture for developers and contributors.
This document provides an overview of the codewhale architecture for developers and contributors.
Current boundary note (v0.8.6):
- `crates/tui` is still the live end-user runtime for the TUI, runtime API, task manager, and tool execution loop.
@@ -178,7 +178,7 @@ drives turns through Chat Completions.
- **`prompts.rs`** - System prompt templates
- **`project_doc.rs`** - Project documentation handling
- **`session.rs`** - Session serialization
- **`runtime_api.rs`** - HTTP/SSE runtime API (`deepseek serve --http`)
- **`runtime_api.rs`** - HTTP/SSE runtime API (`codewhale serve --http`)
- **`runtime_threads.rs`** - Durable thread/turn/item store + replayable event timeline
- **`task_manager.rs`** - Durable queue, worker pool, task timelines and artifacts
docs/CNB_MIRROR.md
+11 -11
View File
@@ -38,9 +38,9 @@ mirror carry them to CNB.
When CNB receives a `v*` tag, the root `.cnb.yml` tag pipeline builds Linux x64
release assets from source and publishes a CNB release with:
- `deepseek-linux-x64`
- `deepseek-tui-linux-x64`
- `deepseek-artifacts-sha256.txt`
- `codewhale-linux-x64`
- `codewhale-tui-linux-x64`
- `codewhale-artifacts-sha256.txt`
This gives users who can reach CNB but not GitHub a CNB-native release path.
GitHub remains the canonical full release matrix; the CNB tag pipeline is the
@@ -144,7 +144,7 @@ expired:
```
4. Confirm the run succeeds via `gh run list --workflow=sync-cnb.yml`.
## Binary release assets and `deepseek update`
## Binary release assets and `codewhale update`
CNB now builds Linux x64 assets for `v*` tags from the source-controlled
`.cnb.yml` pipeline. GitHub remains the canonical full release matrix. Users
@@ -152,24 +152,24 @@ behind GitHub-blocking networks should use one of these paths:
- **`cargo install`** from the CNB mirror:
```bash
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z deepseek-tui-cli
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z deepseek-tui
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z codewhale-cli
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z codewhale-tui
```
(Both binaries are required — the dispatcher and the TUI ship
separately; see `AGENTS.md` for the two-binary install rationale.)
- **CNB release assets** for Linux x64, when the matching CNB tag pipeline has
completed successfully. Download `deepseek-linux-x64`,
`deepseek-tui-linux-x64`, and `deepseek-artifacts-sha256.txt` from the CNB
completed successfully. Download `codewhale-linux-x64`,
`codewhale-tui-linux-x64`, and `codewhale-artifacts-sha256.txt` from the CNB
release for `vX.Y.Z`, then verify the binaries against the manifest.
- **`DEEPSEEK_TUI_RELEASE_BASE_URL`** environment variable, if a
CDN mirror of release assets exists. The npm
wrapper installer and `deepseek update` read this variable to redirect
binary downloads. For `deepseek update`, also set
wrapper installer and `codewhale update` read this variable to redirect
binary downloads. For `codewhale update`, also set
`DEEPSEEK_TUI_VERSION=X.Y.Z` so the updater can label the mirrored
release without contacting GitHub. The directory pointed to must contain
`deepseek-artifacts-sha256.txt` and the platform binaries; format matches
`codewhale-artifacts-sha256.txt` and the platform binaries; format matches
a GitHub Release asset directory.
## Tencent Cloud remote-first path
docs/CONFIGURATION.md
+40 -40
View File
@@ -1,6 +1,6 @@
# Configuration
DeepSeek TUI reads configuration from a TOML file plus environment variables.
codewhale reads configuration from a TOML file plus environment variables.
At process startup it also loads a workspace-local `.env` file when present.
Use the tracked `.env.example` as the template; copy it to `.env`, then edit
only the provider and safety knobs you need.
@@ -13,7 +13,7 @@ Default config path:
Overrides:
- CLI: `deepseek --config /path/to/config.toml`
- CLI: `codewhale --config /path/to/config.toml`
- Env: `DEEPSEEK_CONFIG_PATH=/path/to/config.toml`
If both are set, `--config` wins. Environment variable overrides are applied after the file is loaded.
@@ -49,28 +49,28 @@ Other settings (skills_dir, hooks, capacity, retry, etc.) stay
user-global. If your repo needs more, file an issue describing the
specific use case.
The `deepseek` facade and `deepseek-tui` binary share the same config file for
DeepSeek auth and model defaults. `deepseek auth set --provider deepseek` (and
the legacy `deepseek login --api-key ...` alias) saves the key to
`~/.deepseek/config.toml`, and `deepseek --model deepseek-v4-flash` is forwarded
The `codewhale` facade and `codewhale-tui` binary share the same config file for
DeepSeek auth and model defaults. `codewhale auth set --provider deepseek` (and
the legacy `codewhale login --api-key ...` alias) saves the key to
`~/.deepseek/config.toml`, and `codewhale --model deepseek-v4-flash` is forwarded
to the TUI as `DEEPSEEK_MODEL`.
Credential lookup uses `config -> keyring -> env` after any explicit CLI
`--api-key`. Run `deepseek auth status` to inspect the active provider's config
`--api-key`. Run `codewhale auth status` to inspect the active provider's config
file, OS keyring backend, environment variable, winning source, and last-four
label without printing the key itself. The command only probes the active
provider's keyring entry.
For hosted, generic OpenAI-compatible, or self-hosted providers, set
`provider = "nvidia-nim"`, `"openai"`, `"atlascloud"`, `"wanjie-ark"`, `"fireworks"`,
`"sglang"`, `"vllm"`, or `"ollama"` or pass `deepseek --provider <name>`. The facade saves provider
`"sglang"`, `"vllm"`, or `"ollama"` or pass `codewhale --provider <name>`. The facade saves provider
credentials to the shared user config and forwards the resolved key, base URL,
provider, and model to the TUI process. Use
`deepseek auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"` or
`deepseek auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"` or
`deepseek auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"` or
`deepseek auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"` or
`deepseek auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"` to
`codewhale auth set --provider nvidia-nim --api-key "YOUR_NVIDIA_API_KEY"` or
`codewhale auth set --provider openai --api-key "YOUR_OPENAI_COMPATIBLE_API_KEY"` or
`codewhale auth set --provider atlascloud --api-key "YOUR_ATLASCLOUD_API_KEY"` or
`codewhale auth set --provider wanjie-ark --api-key "YOUR_WANJIE_API_KEY"` or
`codewhale auth set --provider fireworks --api-key "YOUR_FIREWORKS_API_KEY"` to
save provider keys through the facade. The generic `openai` provider defaults
to `https://api.openai.com/v1`, accepts `OPENAI_BASE_URL`, and passes model IDs
through unchanged for OpenAI-compatible gateways. `atlascloud` defaults to
@@ -81,7 +81,7 @@ Wanjie Ark's OpenAI-compatible endpoint at
and passes model IDs through unchanged because Wanjie model access is
account-scoped. SGLang, vLLM, and Ollama are
self-hosted and can run without an API key by default. Ollama defaults to
`http://localhost:11434/v1` and sends model tags such as `deepseek-coder:1.3b`
`http://localhost:11434/v1` and sends model tags such as `codewhale-coder:1.3b`
or `qwen2.5-coder:7b` unchanged. Self-hosted providers and loopback custom
URLs (`localhost`, `127.0.0.1`, `[::1]`, `0.0.0.0`) do not read the secret store
unless API-key auth is explicitly requested; use an env var or config-file key
@@ -113,27 +113,27 @@ when they use localhost or loopback addresses. For a non-local `http://`
gateway, launch with `DEEPSEEK_ALLOW_INSECURE_HTTP=1` only on a trusted network:
```bash
DEEPSEEK_ALLOW_INSECURE_HTTP=1 deepseek
DEEPSEEK_ALLOW_INSECURE_HTTP=1 codewhale
```
Third-party OpenAI-compatible gateways that need extra request headers can set
`http_headers = { "X-Model-Provider-Id" = "your-model-provider" }` at the top
level or under a provider table such as `[providers.deepseek]`. When configured,
DeepSeek TUI sends those custom headers on model API requests. The equivalent
codewhale sends those custom headers on model API requests. The equivalent
environment override is `DEEPSEEK_HTTP_HEADERS`, using comma-separated
`name=value` pairs such as
`X-Model-Provider-Id=your-model-provider,X-Gateway-Route=dev`. `Authorization`
and `Content-Type` are managed by the client and are not overridden by this
setting.
To bootstrap MCP and skills directories at their resolved paths, run `deepseek-tui setup`.
To only scaffold MCP, run `deepseek-tui mcp init`.
To bootstrap MCP and skills directories at their resolved paths, run `codewhale-tui setup`.
To only scaffold MCP, run `codewhale-tui mcp init`.
Note: setup, doctor, mcp, features, sessions, resume/fork, exec, review, and eval
are subcommands of the `deepseek-tui` binary. The `deepseek` dispatcher exposes a
are subcommands of the `codewhale-tui` binary. The `codewhale` dispatcher exposes a
distinct set of commands (`auth`, `config`, `model`, `thread`, `sandbox`,
`app-server`, `mcp-server`, `completion`) and forwards plain prompts to
`deepseek-tui`.
`codewhale-tui`.
## Profiles
@@ -184,15 +184,15 @@ default_text_model = "deepseek-ai/DeepSeek-V4-Pro"
[profiles.ollama]
provider = "ollama"
base_url = "http://localhost:11434/v1"
default_text_model = "deepseek-coder:1.3b"
default_text_model = "codewhale-coder:1.3b"
```
Select a profile with:
- CLI: `deepseek --profile work`
- CLI: `codewhale --profile work`
- Env: `DEEPSEEK_PROFILE=work`
If a profile is selected but missing, DeepSeek TUI exits with an error listing available profiles.
If a profile is selected but missing, codewhale exits with an error listing available profiles.
## Environment Variables
@@ -202,7 +202,7 @@ fallbacks after saved config and keyring credentials:
- `DEEPSEEK_API_KEY`
- `DEEPSEEK_BASE_URL`
- `DEEPSEEK_HTTP_HEADERS` (custom model request headers, comma-separated `name=value` pairs)
- `DEEPSEEK_PROVIDER` (`deepseek|nvidia-nim|openai|atlascloud|wanjie-ark|openrouter|novita|fireworks|sglang|vllm|ollama`)
- `DEEPSEEK_PROVIDER` (`codewhale|nvidia-nim|openai|atlascloud|wanjie-ark|openrouter|novita|fireworks|sglang|vllm|ollama`)
- `DEEPSEEK_MODEL` or `DEEPSEEK_DEFAULT_TEXT_MODEL`
- `DEEPSEEK_STREAM_IDLE_TIMEOUT_SECS` (stream idle timeout in seconds; default `300`, clamped to `1..=3600`)
- `DEEPSEEK_STREAM_OPEN_TIMEOUT_SECS` (connection setup + response-header wait in seconds; default `45`, clamped to `5..=300`; distinct from the per-chunk idle timeout)
@@ -322,7 +322,7 @@ round-trip intact.
## Settings File (Persistent UI Preferences)
DeepSeek TUI also stores user preferences in:
codewhale also stores user preferences in:
- `~/.config/deepseek/settings.toml`
@@ -413,7 +413,7 @@ and the capacity controller remains disabled unless configured.
If you are upgrading from older releases:
- Old: `/deepseek`
- Old: `/codewhale`
New: `/links` (aliases: `/dashboard`, `/api`)
- Old: `/set model deepseek-reasoner`
New: `/config` and edit the `model` row to `deepseek-v4-pro` or `deepseek-v4-flash`
@@ -426,10 +426,10 @@ If you are upgrading from older releases:
### Core keys (used by the TUI/engine)
- `provider` (string, optional): `deepseek` (default), `nvidia-nim`, `openai`, `atlascloud`, `wanjie-ark`, `openrouter`, `novita`, `fireworks`, `sglang`, `vllm`, or `ollama`. Legacy `deepseek-cn` configs are still accepted as an alias for `deepseek`; DeepSeek uses the same official host [`https://api.deepseek.com`](https://api-docs.deepseek.com/) worldwide. `nvidia-nim` targets NVIDIA's NIM-hosted DeepSeek endpoints through `https://integrate.api.nvidia.com/v1`; `openai` targets a generic OpenAI-compatible endpoint, defaulting to `https://api.openai.com/v1`; `atlascloud` targets AtlasCloud's OpenAI-compatible endpoint at `https://api.atlascloud.ai/v1`; `wanjie-ark` targets Wanjie Ark's OpenAI-compatible endpoint at `https://maas-openapi.wanjiedata.com/api/v1`; `fireworks` targets `https://api.fireworks.ai/inference/v1`; `sglang` targets a self-hosted OpenAI-compatible endpoint, defaulting to `http://localhost:30000/v1`; `vllm` targets a self-hosted vLLM OpenAI-compatible endpoint, defaulting to `http://localhost:8000/v1`; `ollama` targets Ollama's OpenAI-compatible endpoint, defaulting to `http://localhost:11434/v1`.
- `provider` (string, optional): `codewhale` (default), `nvidia-nim`, `openai`, `atlascloud`, `wanjie-ark`, `openrouter`, `novita`, `fireworks`, `sglang`, `vllm`, or `ollama`. Legacy `deepseek-cn` configs are still accepted as an alias for `codewhale`; DeepSeek uses the same official host [`https://api.deepseek.com`](https://api-docs.deepseek.com/) worldwide. `nvidia-nim` targets NVIDIA's NIM-hosted DeepSeek endpoints through `https://integrate.api.nvidia.com/v1`; `openai` targets a generic OpenAI-compatible endpoint, defaulting to `https://api.openai.com/v1`; `atlascloud` targets AtlasCloud's OpenAI-compatible endpoint at `https://api.atlascloud.ai/v1`; `wanjie-ark` targets Wanjie Ark's OpenAI-compatible endpoint at `https://maas-openapi.wanjiedata.com/api/v1`; `fireworks` targets `https://api.fireworks.ai/inference/v1`; `sglang` targets a self-hosted OpenAI-compatible endpoint, defaulting to `http://localhost:30000/v1`; `vllm` targets a self-hosted vLLM OpenAI-compatible endpoint, defaulting to `http://localhost:8000/v1`; `ollama` targets Ollama's OpenAI-compatible endpoint, defaulting to `http://localhost:11434/v1`.
- `api_key` (string, required for hosted providers): must be non-empty for DeepSeek/hosted providers (or set the provider API key env var). Self-hosted SGLang, vLLM, and Ollama can omit it.
- `base_url` (string, optional): defaults to `https://api.deepseek.com/beta` for DeepSeek's OpenAI-compatible Chat Completions API, including legacy `provider = "deepseek-cn"` configs, `https://api.openai.com/v1` for `provider = "openai"`, `https://api.atlascloud.ai/v1` for `provider = "atlascloud"`, `https://maas-openapi.wanjiedata.com/api/v1` for `provider = "wanjie-ark"`, or the provider-specific endpoint for hosted/self-hosted providers. Set `https://api.deepseek.com` or `https://api.deepseek.com/v1` explicitly to opt out of DeepSeek beta features.
- `default_text_model` (string, optional): defaults to `deepseek-v4-pro` for DeepSeek, `deepseek-ai/deepseek-v4-pro` for NVIDIA NIM, `gpt-4.1` for generic OpenAI-compatible endpoints, `deepseek-ai/deepseek-v4-flash` for AtlasCloud, `deepseek-reasoner` for Wanjie Ark, `accounts/fireworks/models/deepseek-v4-pro` for Fireworks, `deepseek-ai/DeepSeek-V4-Pro` for SGLang/vLLM, and `deepseek-coder:1.3b` for Ollama. Current public DeepSeek IDs are `deepseek-v4-pro` and `deepseek-v4-flash`, both with 1M context windows, 384K max output, and thinking mode enabled by default. Legacy `deepseek-chat` and `deepseek-reasoner` remain compatibility aliases for `deepseek-v4-flash` until July 24, 2026. Provider-specific mappings translate `deepseek-v4-pro` / `deepseek-v4-flash` to each provider's model ID where supported. Generic `openai`, `atlascloud`, `wanjie-ark`, and Ollama model IDs are passed through unchanged. OpenRouter provider configs with a custom `base_url` also preserve explicit model values, which lets OpenAI-compatible gateways accept bare model IDs. Use `/models` or `deepseek models` to discover live IDs from your configured endpoint. `DEEPSEEK_MODEL` overrides this for a single process.
- `default_text_model` (string, optional): defaults to `deepseek-v4-pro` for DeepSeek, `deepseek-ai/deepseek-v4-pro` for NVIDIA NIM, `gpt-4.1` for generic OpenAI-compatible endpoints, `deepseek-ai/deepseek-v4-flash` for AtlasCloud, `deepseek-reasoner` for Wanjie Ark, `accounts/fireworks/models/deepseek-v4-pro` for Fireworks, `deepseek-ai/DeepSeek-V4-Pro` for SGLang/vLLM, and `codewhale-coder:1.3b` for Ollama. Current public DeepSeek IDs are `deepseek-v4-pro` and `deepseek-v4-flash`, both with 1M context windows, 384K max output, and thinking mode enabled by default. Legacy `deepseek-chat` and `deepseek-reasoner` remain compatibility aliases for `deepseek-v4-flash` until July 24, 2026. Provider-specific mappings translate `deepseek-v4-pro` / `deepseek-v4-flash` to each provider's model ID where supported. Generic `openai`, `atlascloud`, `wanjie-ark`, and Ollama model IDs are passed through unchanged. OpenRouter provider configs with a custom `base_url` also preserve explicit model values, which lets OpenAI-compatible gateways accept bare model IDs. Use `/models` or `codewhale models` to discover live IDs from your configured endpoint. `DEEPSEEK_MODEL` overrides this for a single process.
- `reasoning_effort` (string, optional): `off`, `low`, `medium`, `high`, or `max`; defaults to the configured UI tier. DeepSeek Platform receives top-level `thinking` / `reasoning_effort` fields. NVIDIA NIM receives equivalent settings through `chat_template_kwargs`.
- `allow_shell` (bool, optional): defaults to `true` (sandboxed).
- `approval_policy` (string, optional): `on-request`, `untrusted`, or `never`. Runtime `approval_mode` editing in `/config` also accepts `on-request` and `untrusted` aliases.
@@ -625,10 +625,10 @@ exec_policy = true
You can also override features for a single run:
- `deepseek-tui --enable web_search`
- `deepseek-tui --disable subagents`
- `codewhale-tui --enable web_search`
- `codewhale-tui --disable subagents`
Use `deepseek-tui features list` to inspect known flags and their effective state.
Use `codewhale-tui features list` to inspect known flags and their effective state.
## Web Search Provider
@@ -655,7 +655,7 @@ the composer, press `↑` to select an attachment row, then press `Backspace` or
## Managed Configuration and Requirements
DeepSeek TUI supports a policy layering model:
codewhale supports a policy layering model:
1. user config + profile + env overrides
2. managed config (if present)
@@ -676,17 +676,17 @@ If configured values violate requirements, startup fails with a descriptive erro
See `docs/capacity_controller.md` for formulas, intervention behavior, and telemetry.
## Notes On `deepseek-tui doctor`
## Notes On `codewhale-tui doctor`
`deepseek-tui doctor` follows the same config resolution rules as the rest of the
`codewhale-tui doctor` follows the same config resolution rules as the rest of the
TUI. That means `--config` / `DEEPSEEK_CONFIG_PATH` are respected, and MCP/skills
checks use the resolved `mcp_config_path` / `skills_dir` (including env overrides).
To bootstrap missing MCP/skills paths, run `deepseek-tui setup --all`. You can
also run `deepseek-tui setup --skills --local` to create a workspace-local
To bootstrap missing MCP/skills paths, run `codewhale-tui setup --all`. You can
also run `codewhale-tui setup --skills --local` to create a workspace-local
`./skills` dir.
`deepseek-tui doctor --json` prints a machine-readable report that skips the
`codewhale-tui doctor --json` prints a machine-readable report that skips the
live API connectivity probe. Top-level keys: `version`, `config_path`,
`config_present`, `workspace`, `api_key.source`, `base_url`,
`default_text_model`, `mcp`, `skills`, `tools`, `plugins`, `sandbox`,
@@ -707,7 +707,7 @@ configure reasoning effort.
## Setup status, clean, and extension dirs
`deepseek-tui setup` accepts a few flags beyond the existing `--mcp`,
`codewhale-tui setup` accepts a few flags beyond the existing `--mcp`,
`--skills`, `--local`, `--all`, and `--force`:
- `--status` — print a compact one-screen status (api key, base URL, model,
@@ -732,10 +732,10 @@ configure reasoning effort.
## Why the engine strips XML/`[TOOL_CALL]` text
DeepSeek TUI sends and receives tool calls only over the API tool channel
codewhale sends and receives tool calls only over the API tool channel
(structured `tool_use` / `tool_call` items). The streaming loop in
`crates/tui/src/core/engine.rs` recognizes a fixed set of fake-wrapper start
markers — `[TOOL_CALL]`, `<deepseek:tool_call`, `<tool_call`, `<invoke `,
markers — `[TOOL_CALL]`, `<codewhale:tool_call`, `<tool_call`, `<invoke `,
`<function_calls>` — and scrubs them from visible assistant text without ever
turning them into structured tool calls. When a wrapper is stripped, the loop
emits one compact `status` notice per turn so the user can see why their
docs/DOCKER.md
+13 -13
View File
@@ -12,11 +12,11 @@ docker pull ghcr.io/hmbown/deepseek-tui:latest
Run the published image with a Docker-managed data volume:
```bash
docker volume create deepseek-tui-home
docker volume create codewhale-tui-home
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:latest
@@ -27,7 +27,7 @@ Use a pinned release tag for reproducible installs:
```bash
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
ghcr.io/hmbown/deepseek-tui:vX.Y.Z
@@ -41,7 +41,7 @@ Replace `vX.Y.Z` with a tag from
Build the image locally from a checkout:
```bash
docker build -t deepseek-tui .
docker build -t codewhale-tui .
```
Then run it with the same Docker-managed data volume:
@@ -49,10 +49,10 @@ Then run it with the same Docker-managed data volume:
```bash
docker run --rm -it \
-e DEEPSEEK_API_KEY="$DEEPSEEK_API_KEY" \
-v deepseek-tui-home:/home/deepseek/.deepseek \
-v codewhale-tui-home:/home/deepseek/.deepseek \
-v "$PWD:/workspace" \
-w /workspace \
deepseek-tui
codewhale-tui
```
Docker Hub publishing is not configured; GHCR is the supported prebuilt image
@@ -74,13 +74,13 @@ is the safest default because Docker creates it with ownership the container can
write:
```bash
-v deepseek-tui-home:/home/deepseek/.deepseek
-v codewhale-tui-home:/home/deepseek/.deepseek
```
Without this mount the container starts fresh each time.
If you bind-mount an existing host directory instead, the image runs as the
non-root `deepseek` user with UID/GID `1000:1000`. The mounted directory must be
non-root `codewhale` user with UID/GID `1000:1000`. The mounted directory must be
writable by that user, or startup can fail while creating runtime directories
under `.deepseek/tasks`. On Linux hosts, either use the named volume above or
prepare the bind mount explicitly:
@@ -101,8 +101,8 @@ volume instead.
## Non-interactive / pipeline usage
When stdin is not a TTY, `deepseek` drops to the dispatcher's one-shot mode
(`deepseek -c "…"`). Pipe a prompt on stdin:
When stdin is not a TTY, `codewhale` drops to the dispatcher's one-shot mode
(`codewhale -c "…"`). Pipe a prompt on stdin:
```bash
echo "Explain the Cargo.toml in structured English." | \
@@ -113,18 +113,18 @@ echo "Explain the Cargo.toml in structured English." | \
```bash
# Single platform (your host architecture)
docker build -t deepseek-tui .
docker build -t codewhale-tui .
# Multi-platform (requires a builder with emulation)
docker buildx create --use
docker buildx build --platform linux/amd64,linux/arm64 -t deepseek-tui .
docker buildx build --platform linux/amd64,linux/arm64 -t codewhale-tui .
```
## Devcontainer
The repository includes a [`.devcontainer/devcontainer.json`](../.devcontainer/devcontainer.json)
configuration for VS Code / GitHub Codespaces. It pre-installs the Rust toolchain,
rust-analyzer, and the `deepseek` binary. Open the repo in a devcontainer to get a
rust-analyzer, and the `codewhale` binary. Open the repo in a devcontainer to get a
ready-to-use development environment.
## Release status
docs/INSTALL.md
+77 -77
View File
@@ -1,4 +1,4 @@
# Installing DeepSeek TUI
# Installing codewhale
This page covers every supported install path and the most common
"it didn't install" failures, including **Linux ARM64** and other less
@@ -12,16 +12,16 @@ If you just want the short version, see the
## 1. Supported platforms
`deepseek-tui` ships prebuilt binaries for these
`codewhale-tui` ships prebuilt binaries for these
platform/architecture combinations from v0.8.8 onward:
| Platform | Architecture | npm install | `cargo install` | GitHub release asset |
| ------------ | ------------ | :---------: | :-------------: | ----------------------------------------------------- |
| Linux | x64 (x86_64) | ✅ | ✅ | `deepseek-linux-x64`, `deepseek-tui-linux-x64` |
| Linux | arm64 | ✅ | ✅ | `deepseek-linux-arm64`, `deepseek-tui-linux-arm64` |
| macOS | x64 | ✅ | ✅ | `deepseek-macos-x64`, `deepseek-tui-macos-x64` |
| macOS | arm64 (M-series) | ✅ | ✅ | `deepseek-macos-arm64`, `deepseek-tui-macos-arm64` |
| Windows | x64 | ✅ | ✅ | `deepseek-windows-x64.exe`, `deepseek-tui-windows-x64.exe` |
| Linux | x64 (x86_64) | ✅ | ✅ | `codewhale-linux-x64`, `codewhale-tui-linux-x64` |
| Linux | arm64 | ✅ | ✅ | `codewhale-linux-arm64`, `codewhale-tui-linux-arm64` |
| macOS | x64 | ✅ | ✅ | `codewhale-macos-x64`, `codewhale-tui-macos-x64` |
| macOS | arm64 (M-series) | ✅ | ✅ | `codewhale-macos-arm64`, `codewhale-tui-macos-arm64` |
| Windows | x64 | ✅ | ✅ | `codewhale-windows-x64.exe`, `codewhale-tui-windows-x64.exe` |
| Other Linux (musl, riscv64, …) | — | ❌¹ | ✅² | build from source |
| FreeBSD / OpenBSD | — | ❌ | ✅² | build from source |
@@ -38,8 +38,8 @@ systems such as Alpine should use [Build from source](#7-build-from-source).
> **Linux ARM64 note (v0.8.7 and earlier).** v0.8.7 and earlier do **not**
> publish a Linux ARM64 prebuilt; users on HarmonyOS thin-and-light, Asahi
> Linux, Raspberry Pi, AWS Graviton, etc. saw `Unsupported architecture: arm64`
> from `npm i -g deepseek-tui`. v0.8.8 publishes both `deepseek-linux-arm64`
> and `deepseek-tui-linux-arm64`, so a plain `npm i -g deepseek-tui` works
> from `npm i -g codewhale`. v0.8.8 publishes both `codewhale-linux-arm64`
> and `codewhale-tui-linux-arm64`, so a plain `npm i -g codewhale` works
> on any glibc-based ARM64 Linux. If you're stuck on v0.8.7, jump to
> [Build from source](#7-build-from-source) — `cargo install` works fine.
@@ -49,19 +49,19 @@ systems such as Alpine should use [Build from source](#7-build-from-source).
Official release binaries are published only from
`https://github.com/Hmbown/DeepSeek-TUI/releases` and the npm package named
`deepseek-tui`. Do not install release assets from look-alike repositories,
`codewhale-tui`. Do not install release assets from look-alike repositories,
archives, or search-result mirrors unless you deliberately trust that mirror.
Every GitHub release includes `deepseek-artifacts-sha256.txt`. If you download
Every GitHub release includes `codewhale-artifacts-sha256.txt`. If you download
binaries manually, verify them before running:
```bash
# Run from the directory containing the downloaded binaries.
curl -L -O https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/deepseek-artifacts-sha256.txt
sha256sum -c deepseek-artifacts-sha256.txt --ignore-missing
curl -L -O https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/codewhale-artifacts-sha256.txt
sha256sum -c codewhale-artifacts-sha256.txt --ignore-missing
```
On macOS, use `shasum -a 256 -c deepseek-artifacts-sha256.txt` instead of
On macOS, use `shasum -a 256 -c codewhale-artifacts-sha256.txt` instead of
`sha256sum`.
If antivirus software flags an official release binary, treat it as unresolved
@@ -70,7 +70,7 @@ the GitHub issue:
- the release tag, for example `v0.8.36`
- the exact download URL
- the filename, for example `deepseek-linux-x64`
- the filename, for example `codewhale-linux-x64`
- the file SHA-256 from your machine
- the antivirus product name and detection name
@@ -82,13 +82,13 @@ a download sourced from an impersonating repository or mirror.
## 3. Install via npm (recommended)
```bash
npm install -g deepseek-tui
deepseek
npm install -g codewhale
codewhale
```
`postinstall` downloads the right pair of binaries from the matching GitHub
release, verifies a SHA-256 manifest, and exposes both `deepseek` and
`deepseek-tui` on your `PATH`.
release, verifies a SHA-256 manifest, and exposes both `codewhale` and
`codewhale-tui` on your `PATH`.
Useful environment variables:
@@ -105,7 +105,7 @@ Useful environment variables:
> (not just the postinstall binary download), use an npm registry mirror:
> ```bash
> npm config set registry https://registry.npmmirror.com
> npm install -g deepseek-tui
> npm install -g codewhale
> ```
> See also [Section 4](#4-install-via-cargo-any-tier-1-rust-target) if you
> prefer Cargo over npm.
@@ -120,9 +120,9 @@ delegates to the TUI runtime at runtime.
```bash
# Requires Rust 1.88+ (https://rustup.rs)
cargo install deepseek-tui-cli --locked # provides `deepseek`
cargo install deepseek-tui --locked # provides `deepseek-tui`
deepseek --version
cargo install codewhale-cli --locked # provides `codewhale`
cargo install codewhale-tui --locked # provides `codewhale-tui`
codewhale --version
```
### China / mirror-friendly install
@@ -201,7 +201,7 @@ If you already have Nix with flake support, run:
nix run github:Hmbown/DeepSeek-TUI
```
Nix builds `deepseek-tui` and then starts the `deepseek` dispatcher. Pass
Nix builds `codewhale-tui` and then starts the `codewhale` dispatcher. Pass
arguments after `--`, for example:
```sh
@@ -217,8 +217,8 @@ Add inputs to `flake.nix`:
inputs = {
nixpkgs.url = "github:NixOS/nixpkgs/nixos-unstable";
deepseek-tui.url = "github:Hmbown/DeepSeek-TUI";
deepseek-tui.inputs.nixpkgs.follows = "nixpkgs";
codewhale-tui.url = "github:Hmbown/DeepSeek-TUI";
codewhale-tui.inputs.nixpkgs.follows = "nixpkgs";
};
}
```
@@ -227,7 +227,7 @@ Install into a NixOS module:
```nix
{
outputs = { self, nixpkgs, deepseek-tui }:
outputs = { self, nixpkgs, codewhale-tui }:
let
# replace system "x86_64-linux" with your system
system = "x86_64-linux";
@@ -239,7 +239,7 @@ Install into a NixOS module:
modules = [
# ...
{
environment.systemPackages = [ deepseek-tui.packages.${system}.default ];
environment.systemPackages = [ codewhale-tui.packages.${system}.default ];
}
];
};
@@ -258,32 +258,32 @@ side by side into a directory on your `PATH` (e.g. `~/.local/bin`):
```bash
# Linux ARM64 example
mkdir -p ~/.local/bin
curl -L -o ~/.local/bin/deepseek \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/deepseek-linux-arm64
curl -L -o ~/.local/bin/deepseek-tui \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/deepseek-tui-linux-arm64
chmod +x ~/.local/bin/deepseek ~/.local/bin/deepseek-tui
deepseek --version
curl -L -o ~/.local/bin/codewhale \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/codewhale-linux-arm64
curl -L -o ~/.local/bin/codewhale-tui \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/codewhale-tui-linux-arm64
chmod +x ~/.local/bin/codewhale ~/.local/bin/codewhale-tui
codewhale --version
```
Verify integrity against the per-release SHA-256 manifest:
```bash
curl -L -o /tmp/deepseek-artifacts-sha256.txt \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/deepseek-artifacts-sha256.txt
( cd ~/.local/bin && sha256sum -c /tmp/deepseek-artifacts-sha256.txt --ignore-missing )
curl -L -o /tmp/codewhale-artifacts-sha256.txt \
https://github.com/Hmbown/DeepSeek-TUI/releases/latest/download/codewhale-artifacts-sha256.txt
( cd ~/.local/bin && sha256sum -c /tmp/codewhale-artifacts-sha256.txt --ignore-missing )
```
(Use `shasum -a 256 -c` instead of `sha256sum` on macOS.)
### Windows Scoop
DeepSeek TUI is listed in Scoop's main bucket:
codewhale is listed in Scoop's main bucket:
```powershell
scoop update
scoop install deepseek-tui
deepseek --version
codewhale --version
```
Scoop manifests are maintained outside this repository's release workflow and
@@ -314,10 +314,10 @@ LoongArch, FreeBSD, and pre-2024 ARM64 distros.
git clone https://github.com/Hmbown/DeepSeek-TUI.git
cd DeepSeek-TUI
cargo install --path crates/cli --locked # provides `deepseek`
cargo install --path crates/tui --locked # provides `deepseek-tui`
cargo install --path crates/cli --locked # provides `codewhale`
cargo install --path crates/tui --locked # provides `codewhale-tui`
deepseek --version
codewhale --version
```
Both binaries land in `~/.cargo/bin/` by default; make sure that directory is
@@ -336,13 +336,13 @@ rustup target add aarch64-unknown-linux-gnu
cargo install cross --locked
# Per build
cross build --release --target aarch64-unknown-linux-gnu -p deepseek-tui-cli
cross build --release --target aarch64-unknown-linux-gnu -p deepseek-tui
cross build --release --target aarch64-unknown-linux-gnu -p codewhale-cli
cross build --release --target aarch64-unknown-linux-gnu -p codewhale-tui
```
The resulting binaries land in
`target/aarch64-unknown-linux-gnu/release/deepseek` and
`target/aarch64-unknown-linux-gnu/release/deepseek-tui`. Copy the matched pair
`target/aarch64-unknown-linux-gnu/release/codewhale` and
`target/aarch64-unknown-linux-gnu/release/codewhale-tui`. Copy the matched pair
to the ARM64 host (e.g. via `scp`) and `chmod +x` them.
If you don't have Docker available, install the cross-linker directly and let
@@ -357,8 +357,8 @@ cat >> ~/.cargo/config.toml <<'EOF'
linker = "aarch64-linux-gnu-gcc"
EOF
cargo build --release --target aarch64-unknown-linux-gnu -p deepseek-tui-cli
cargo build --release --target aarch64-unknown-linux-gnu -p deepseek-tui
cargo build --release --target aarch64-unknown-linux-gnu -p codewhale-cli
cargo build --release --target aarch64-unknown-linux-gnu -p codewhale-tui
```
The same recipe works for `aarch64-unknown-linux-musl` if your distro is
@@ -420,8 +420,8 @@ set CARGO_HTTP_CHECK_REVOKE=false # may be needed behind some Chinese ISPs
cargo build --release
```
Both binaries appear in `target\release\deepseek.exe` and
`target\release\deepseek-tui.exe`.
Both binaries appear in `target\release\codewhale.exe` and
`target\release\codewhale-tui.exe`.
> **Prefer `npm install -g` on Windows unless you need to modify source.**
> The npm package pulls prebuilt binaries and avoids the C toolchain
@@ -434,30 +434,30 @@ Both binaries appear in `target\release\deepseek.exe` and
### `Unsupported architecture: arm64 on platform linux`
You're on a release earlier than v0.8.8 that doesn't publish Linux ARM64
binaries. Either upgrade (`npm i -g deepseek-tui@latest`) or use
binaries. Either upgrade (`npm i -g codewhale@latest`) or use
`cargo install` per [Section 4](#4-install-via-cargo-any-tier-1-rust-target).
### `MISSING_COMPANION_BINARY` at runtime
The dispatcher (`deepseek`) requires the TUI runtime (`deepseek-tui`) to be on
The dispatcher (`codewhale`) requires the TUI runtime (`codewhale-tui`) to be on
the same `PATH`. If you installed only one crate via `cargo install`, install
both:
```bash
cargo install deepseek-tui-cli --locked
cargo install deepseek-tui --locked
cargo install codewhale-cli --locked
cargo install codewhale-tui --locked
```
### `deepseek update` reports `no asset found for platform deepseek-linux-aarch64`
### `codewhale update` reports `no asset found for platform codewhale-linux-aarch64`
This is [#503](https://github.com/Hmbown/DeepSeek-TUI/issues/503) in v0.8.7 —
the self-updater used Rust's `aarch64`/`x86_64` arch names instead of the
release artifact's `arm64`/`x64`. Workaround until v0.8.8:
```bash
npm i -g deepseek-tui@latest
npm i -g codewhale@latest
# or
cargo install deepseek-tui-cli --locked
cargo install codewhale-cli --locked
```
### npm download is slow or times out from mainland China
@@ -466,26 +466,26 @@ Set `DEEPSEEK_TUI_RELEASE_BASE_URL` to a mirrored release-asset directory
(rsproxy, TUNA, Tencent COS, Aliyun OSS), or skip npm entirely and use the
Cargo mirror setup in [Section 4](#4-install-via-cargo-any-tier-1-rust-target).
### `deepseek update` is blocked by GitHub from mainland China
### `codewhale update` is blocked by GitHub from mainland China
`deepseek update` normally contacts GitHub Releases for metadata and binary
`codewhale update` normally contacts GitHub Releases for metadata and binary
assets. On networks where GitHub is blocked or unreliable, use the CNB source
mirror instead and install both binaries from the release tag:
```bash
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z deepseek-tui-cli --locked --force
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z deepseek-tui --locked --force
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z codewhale-cli --locked --force
cargo install --git https://cnb.cool/deepseek-tui.com/DeepSeek-TUI --tag vX.Y.Z codewhale-tui --locked --force
```
If you operate a binary asset mirror, `deepseek update` can use it directly:
If you operate a binary asset mirror, `codewhale update` can use it directly:
```bash
DEEPSEEK_TUI_VERSION=X.Y.Z \
DEEPSEEK_TUI_RELEASE_BASE_URL=https://your-mirror.example.com/DeepSeek-TUI/vX.Y.Z/ \
deepseek update
codewhale update
```
The mirror directory must contain `deepseek-artifacts-sha256.txt` and the
The mirror directory must contain `codewhale-artifacts-sha256.txt` and the
platform binaries from the GitHub release.
### Debian/Ubuntu: `feature edition2024 is required` from `cargo install`
@@ -511,8 +511,8 @@ export RUSTUP_UPDATE_ROOT=https://rsproxy.cn/rustup
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y
source "$HOME/.cargo/env"
rustup default stable
cargo install deepseek-tui-cli --locked
cargo install deepseek-tui --locked
cargo install codewhale-cli --locked
cargo install codewhale-tui --locked
```
Afterward, `which cargo` should point to `~/.cargo/bin/cargo`, not
@@ -526,7 +526,7 @@ Install the C toolchain:
sudo apt-get install -y build-essential pkg-config libdbus-1-dev
```
### Wrapper installs but `deepseek` isn't found
### Wrapper installs but `codewhale` isn't found
`npm i -g` installs into `$(npm prefix -g)/bin`; make sure that directory is on
your shell's `PATH`. With nvm: `nvm use --lts && hash -r`.
@@ -573,10 +573,10 @@ path-agnostic — moving `target-dir` does not help.
1. **Add the project's `target/` directory to your AV exclusions list.**
2. **Close the antivirus software temporarily** during `cargo build`.
3. **Use `npm install -g deepseek-tui` instead** — the npm package ships
3. **Use `npm install -g codewhale` instead** — the npm package ships
prebuilt binaries and skips the Cargo build entirely
([Section 3](#3-install-via-npm-recommended)).
4. **Use `cargo install deepseek-tui-cli --locked`** from crates.io — this
4. **Use `cargo install codewhale-cli --locked`** from crates.io — this
changes the binary path, which some AV tools treat differently.
To verify that the build-script binary itself is valid (not corrupted), locate
@@ -590,7 +590,7 @@ target/debug/build/libsqlite3-sys-*/build-script-build
### npm binary download times out
If `deepseek` waits several seconds and prints `connect ETIMEDOUT` or
If `codewhale` waits several seconds and prints `connect ETIMEDOUT` or
`EAI_AGAIN` while fetching from `github.com`, the npm wrapper installed
successfully but the prebuilt binary download from GitHub Releases is blocked
or unreliable on your network. This download is separate from the npm registry
@@ -602,23 +602,23 @@ Use one of these paths:
```bash
export HTTPS_PROXY=http://your-proxy:port
deepseek
codewhale
```
2. Mirror the release assets internally and set `DEEPSEEK_TUI_RELEASE_BASE_URL`:
```bash
export DEEPSEEK_TUI_RELEASE_BASE_URL=https://your-mirror.example.com/DeepSeek-TUI/
deepseek
codewhale
```
The directory must contain `deepseek-artifacts-sha256.txt` and the platform
The directory must contain `codewhale-artifacts-sha256.txt` and the platform
binaries from the GitHub release.
3. Install via Cargo, which builds locally and does not download GitHub release
assets. See [Section 4](#4-install-via-cargo-any-tier-1-rust-target).
4. Download both `deepseek` and `deepseek-tui` manually from the
4. Download both `codewhale` and `codewhale-tui` manually from the
[Releases page](https://github.com/Hmbown/DeepSeek-TUI/releases), place them
in a directory on `PATH`, and make them executable. See
[Section 6](#6-manual-download-from-github-releases).
@@ -628,9 +628,9 @@ Use one of these paths:
## 9. Verifying your install
```bash
deepseek --version
deepseek doctor # checks API key, provider, runtime, and PATH integrity
deepseek doctor --json
codewhale --version
codewhale doctor # checks API key, provider, runtime, and PATH integrity
codewhale doctor --json
```
`doctor` exits non-zero if it finds a problem and prints structured remediation
docs/MCP.md
+29 -29
View File
@@ -1,15 +1,15 @@
# MCP (External Tool Servers)
DeepSeek TUI can load additional tools via MCP (Model Context Protocol). MCP servers are local processes that the TUI starts and communicates with over stdio.
codewhale can load additional tools via MCP (Model Context Protocol). MCP servers are local processes that the TUI starts and communicates with over stdio.
Browsing note:
- `web.run` is the canonical built-in browsing tool.
- `web_search` remains available as a compatibility alias for older prompts and integrations.
Server mode note:
- `deepseek-tui serve --mcp` runs the MCP stdio server.
- `deepseek-tui serve --http` runs the runtime HTTP/SSE API (separate mode).
- The `deepseek` dispatcher exposes `deepseek mcp-server` as an equivalent stdio
- `codewhale-tui serve --mcp` runs the MCP stdio server.
- `codewhale-tui serve --http` runs the runtime HTTP/SSE API (separate mode).
- The `codewhale` dispatcher exposes `codewhale mcp-server` as an equivalent stdio
entrypoint used by the split CLI.
## Bootstrap MCP Config
@@ -17,22 +17,22 @@ Server mode note:
Create a starter MCP config at your resolved MCP path:
```bash
deepseek-tui mcp init
codewhale-tui mcp init
```
`deepseek-tui setup --mcp` performs the same MCP bootstrap alongside skills setup.
`codewhale-tui setup --mcp` performs the same MCP bootstrap alongside skills setup.
Common management commands:
```bash
deepseek-tui mcp list
deepseek-tui mcp tools [server]
deepseek-tui mcp add <name> --command "<cmd>" --arg "<arg>"
deepseek-tui mcp add <name> --url "http://localhost:3000/mcp"
deepseek-tui mcp enable <name>
deepseek-tui mcp disable <name>
deepseek-tui mcp remove <name>
deepseek-tui mcp validate
codewhale-tui mcp list
codewhale-tui mcp tools [server]
codewhale-tui mcp add <name> --command "<cmd>" --arg "<arg>"
codewhale-tui mcp add <name> --url "http://localhost:3000/mcp"
codewhale-tui mcp enable <name>
codewhale-tui mcp disable <name>
codewhale-tui mcp remove <name>
codewhale-tui mcp validate
```
## In-TUI Manager
@@ -72,7 +72,7 @@ Overrides:
- Config: `mcp_config_path = "/path/to/mcp.json"`
- Env: `DEEPSEEK_MCP_CONFIG=/path/to/mcp.json`
`deepseek-tui mcp init` (and `deepseek-tui setup --mcp`) writes to this resolved path.
`codewhale-tui mcp init` (and `codewhale-tui setup --mcp`) writes to this resolved path.
The interactive `/config` editor also exposes `mcp_config_path`. Changing it in
the TUI updates the path used by `/mcp`, and requires a restart before the
@@ -130,14 +130,14 @@ You can register your local DeepSeek binary as an MCP server so other DeepSeek s
### Quick Setup
```bash
deepseek-tui mcp add-self
codewhale-tui mcp add-self
```
This resolves the current binary path, generates a config entry that runs `deepseek-tui serve --mcp`, and writes it to your MCP config file. The default server name is `deepseek`.
This resolves the current binary path, generates a config entry that runs `codewhale-tui serve --mcp`, and writes it to your MCP config file. The default server name is `codewhale`.
Options:
- `--name <NAME>` — custom server name (default: `deepseek`)
- `--name <NAME>` — custom server name (default: `codewhale`)
- `--workspace <PATH>` — workspace directory for the server
### Manual Config
@@ -147,8 +147,8 @@ Equivalent manual entry in `~/.deepseek/mcp.json`:
```json
{
"servers": {
"deepseek": {
"command": "/path/to/deepseek",
"codewhale": {
"command": "/path/to/codewhale",
"args": ["serve", "--mcp"],
"env": {}
}
@@ -156,9 +156,9 @@ Equivalent manual entry in `~/.deepseek/mcp.json`:
}
```
The `deepseek-tui` binary supports `serve --mcp` directly. The `deepseek`
dispatcher offers the equivalent `deepseek mcp-server` stdio entrypoint. Use
whichever is on your `PATH` (run `which deepseek` or `which deepseek-tui` to
The `codewhale-tui` binary supports `serve --mcp` directly. The `codewhale`
dispatcher offers the equivalent `codewhale mcp-server` stdio entrypoint. Use
whichever is on your `PATH` (run `which codewhale` or `which codewhale-tui` to
find the full path). The `mcp add-self` command automatically resolves the
correct binary.
@@ -172,13 +172,13 @@ correct binary.
Tools from a self-hosted DeepSeek server follow the standard naming convention:
- `mcp_deepseek_<tool>` (if the server is named `deepseek`)
- `mcp_deepseek_<tool>` (if the server is named `codewhale`)
For example, the `shell` tool becomes `mcp_deepseek_shell`.
### MCP Server vs HTTP/SSE API vs ACP
| | `deepseek-tui serve --mcp` | `deepseek-tui serve --http` | `deepseek-tui serve --acp` |
| | `codewhale-tui serve --mcp` | `codewhale-tui serve --http` | `codewhale-tui serve --acp` |
|---|---|---|---|
| **Protocol** | MCP stdio | HTTP/SSE JSON-RPC | ACP stdio |
| **Use case** | Tool server for MCP clients | Runtime API for apps | Editor agent for Zed/custom ACP clients |
@@ -194,8 +194,8 @@ Use `serve --acp` when an editor wants to talk to DeepSeek as an ACP agent.
After adding, test the connection:
```bash
deepseek-tui mcp validate
deepseek-tui mcp tools deepseek
codewhale-tui mcp validate
codewhale-tui mcp tools codewhale
```
## Server Fields
@@ -220,7 +220,7 @@ You should still only configure MCP servers you trust, and treat MCP server conf
## Troubleshooting
- Run `deepseek-tui doctor` to confirm the MCP config path it resolved and whether it exists.
- Run `codewhale-tui doctor` to confirm the MCP config path it resolved and whether it exists.
- In the TUI, run `/mcp validate` to refresh the visible server/tool snapshot.
- If the MCP config is missing, run `deepseek-tui mcp init --force` to regenerate it.
- If the MCP config is missing, run `codewhale-tui mcp init --force` to regenerate it.
- If tools don’t appear, verify the server command works from your shell and that the server supports MCP `tools/list`.
docs/MODES.md
+8 -8
View File
@@ -1,6 +1,6 @@
# Modes and Approvals
DeepSeek TUI has two related concepts:
codewhale has two related concepts:
- **TUI mode**: what kind of visible interaction you're in (Plan/Agent/YOLO).
- **Approval mode**: how aggressively the UI asks before executing tools.
@@ -75,14 +75,14 @@ See `MCP.md`.
## Related CLI Flags
Run `deepseek --help` for the canonical list. Common flags:
Run `codewhale --help` for the canonical list. Common flags:
- `-p, --prompt <TEXT>`: one-shot prompt mode (prints and exits)
- `deepseek exec --output-format stream-json <PROMPT>`: emit one JSON object per line for harnesses and backend wrappers
- `deepseek exec --resume <ID|PREFIX> <PROMPT>` / `--session-id <ID|PREFIX>`: continue a saved session non-interactively
- `deepseek exec --continue <PROMPT>`: continue the most recent saved session for this workspace non-interactively
- `deepseek fork <ID|PREFIX>` / `deepseek fork --last`: copy a saved session into a new sibling session; forked sessions retain additive parent-session metadata and show that lineage in session listings
- `--model <MODEL>`: when using the `deepseek` facade, forward a DeepSeek model override to the TUI
- `codewhale exec --output-format stream-json <PROMPT>`: emit one JSON object per line for harnesses and backend wrappers
- `codewhale exec --resume <ID|PREFIX> <PROMPT>` / `--session-id <ID|PREFIX>`: continue a saved session non-interactively
- `codewhale exec --continue <PROMPT>`: continue the most recent saved session for this workspace non-interactively
- `codewhale fork <ID|PREFIX>` / `codewhale fork --last`: copy a saved session into a new sibling session; forked sessions retain additive parent-session metadata and show that lineage in session listings
- `--model <MODEL>`: when using the `codewhale` facade, forward a DeepSeek model override to the TUI
- `--workspace <DIR>`: workspace root for file tools
- `--yolo`: start in YOLO mode
- `-r, --resume <ID|PREFIX|latest>`: resume a saved session
@@ -97,7 +97,7 @@ Run `deepseek --help` for the canonical list. Common flags:
DeepSeek-TUI has three related but intentionally separate recovery paths:
- `deepseek fork <ID>` creates a new saved session from an existing saved
- `codewhale fork <ID>` creates a new saved session from an existing saved
conversation and records the source session id. This is the safe way to
explore a different answer path without overwriting the original session.
- Esc-Esc backtrack rewinds the live transcript to a previous user prompt and
docs/OPERATIONS_RUNBOOK.md
+2 -2
View File
@@ -1,4 +1,4 @@
# DeepSeek TUI Operations Runbook
# codewhale Operations Runbook
This runbook covers practical debugging and incident response for the local CLI/TUI runtime.
@@ -56,7 +56,7 @@ Expected behavior:
- Startup begins a fresh session unless `--resume`/`--continue` is supplied
Actions:
1. Resume prior work explicitly via `deepseek --resume <id>` or `Ctrl+R` in TUI
1. Resume prior work explicitly via `codewhale --resume <id>` or `Ctrl+R` in TUI
2. If checkpoint inspection is needed, inspect `latest.json` for schema mismatch/details
3. If schema is newer than binary supports, upgrade binary or remove stale checkpoint
docs/REBRAND.md
+138
View File
@@ -0,0 +1,138 @@
# Rebrand: deepseek-tui → codewhale
Starting with **v0.8.41**, this project ships under a new name: `codewhale`.
This document explains what changed, what didn't, and how to migrate. None of the
DeepSeek provider integration changed — only the local CLI / TUI brand.
## TL;DR
```bash
# 1. Uninstall the old wrapper or binaries.
npm uninstall -g deepseek-tui # or cargo uninstall deepseek-tui-cli deepseek-tui
# or brew uninstall deepseek-tui
# 2. Install under the new name.
npm install -g codewhale # or cargo install codewhale-cli codewhale-tui --locked
# or brew install deepseek-tui (Homebrew tap still
# uses the legacy name during the transition;
# it installs the new binaries underneath.)
# 3. Run with the new command.
codewhale doctor
codewhale
```
Your `~/.deepseek/config.toml`, `~/.deepseek/sessions/`, `~/.deepseek/skills/`,
`~/.deepseek/tasks/`, and `~/.deepseek/mcp.json` are untouched. Existing
`DEEPSEEK_*` environment variables continue to work.
## What got renamed
| Surface | Before | After |
|---|---|---|
| CLI dispatcher binary | `deepseek` | `codewhale` |
| TUI runtime binary | `deepseek-tui` | `codewhale-tui` |
| npm wrapper package | `deepseek-tui` | `codewhale` |
| Crates.io crates | `deepseek-tui-cli` / `deepseek-tui` / `deepseek-*` | `codewhale-cli` / `codewhale-tui` / `codewhale-*` |
| Release assets | `deepseek-<platform>` / `deepseek-tui-<platform>` | `codewhale-<platform>` / `codewhale-tui-<platform>` |
| Checksum manifest | `deepseek-artifacts-sha256.txt` | `codewhale-artifacts-sha256.txt` |
## What did NOT change
Anything that targets the DeepSeek provider API stays exactly as it was:
- **Environment variables**: `DEEPSEEK_API_KEY`, `DEEPSEEK_BASE_URL`,
`DEEPSEEK_MODEL`, `DEEPSEEK_PROVIDER`, `DEEPSEEK_PROFILE`, `DEEPSEEK_YOLO`,
`DEEPSEEK_LOG_LEVEL`, plus the existing `DEEPSEEK_TUI_*` runtime knobs
(`DEEPSEEK_TUI_BIN`, `DEEPSEEK_TUI_RELEASE_BASE_URL`, etc.). They're kept
for backward compatibility; renaming them would break every shell rc on
the planet.
- **Model IDs**: `deepseek-v4-pro`, `deepseek-v4-flash`, and the legacy
aliases `deepseek-chat` and `deepseek-reasoner`.
- **Hosts**: `api.deepseek.com` (global) and `api.deepseeki.com` (China
fallback).
- **Config directory**: `~/.deepseek/`. Renaming this would invalidate
every existing install's saved API key, sessions, skills, MCP config,
and audit log.
- **GitHub repository URL**: still `https://github.com/Hmbown/DeepSeek-TUI`.
A repo rename is a maintainer-side operation that ships separately.
- **Homebrew tap and formula** (`Hmbown/homebrew-deepseek-tui`): still
installs by the legacy name during the transition. The tap's formula
will be flipped to the new names in a follow-up.
- **Docker image** (`ghcr.io/hmbown/deepseek-tui`): unchanged.
## Deprecation shims (one release cycle)
To keep existing shell aliases, scripts, and CI working through the rename,
v0.8.41 ships **deprecation shims** for one cycle:
- A `deepseek` binary that prints a one-line warning to stderr and forwards
argv to `codewhale`.
- A `deepseek-tui` binary that does the same for `codewhale-tui`.
- An `npm` package at `deepseek-tui@0.8.41` with no `bin` and a postinstall
that prints a clear "rename" notice.
These shims will be removed in **v0.9.0**. Please migrate before then.
## Migrating in practice
### npm
```bash
npm uninstall -g deepseek-tui
npm install -g codewhale
```
### Cargo
```bash
cargo uninstall deepseek-tui-cli deepseek-tui 2>/dev/null || true
cargo install codewhale-cli codewhale-tui --locked
```
Or in a checkout:
```bash
cargo install --path crates/cli --locked --force
cargo install --path crates/tui --locked --force
```
### Homebrew
The tap formula still installs `deepseek-tui` during the transition.
Existing `brew install deepseek-tui` invocations continue to work and land
the new binaries underneath the legacy formula name. The formula and tap
repo will follow up with their own rename.
### Manual / GitHub Releases
`v0.8.41` Releases attach **both** the canonical `codewhale-*` /
`codewhale-tui-*` assets and the legacy `deepseek-*` / `deepseek-tui-*`
shim assets. Existing `deepseek update` invocations on v0.8.40 keep working;
they land you on the deprecation shim, which then prompts the install of
`codewhale`.
A second checksum manifest, `deepseek-artifacts-sha256.txt`, is attached as
an alias of `codewhale-artifacts-sha256.txt` so v0.8.40's hardcoded lookup
still verifies.
## Why the name change
`codewhale` is a shorter, terminal-friendlier handle that doesn't suggest
the project is tied to a single provider. The dispatcher already supports
DeepSeek, NVIDIA NIM, OpenAI-compatible endpoints, AtlasCloud, Wanjie Ark,
OpenRouter, Novita, Fireworks, SGLang, vLLM, and Ollama, with more on the
roadmap. The new name reflects that.
## Reporting issues with the rename
If your install broke during the migration, please open an issue at
<https://github.com/Hmbown/DeepSeek-TUI/issues> and include:
- The output of `codewhale --version` (or `deepseek --version` if you're
still on the shim).
- Which install path you used (npm, cargo, brew, manual).
- The exact command you ran and the full error output.
We'll prioritize migration regressions.
docs/RELEASE_CHECKLIST.md
+3 -3
View File
@@ -29,7 +29,7 @@ publish-crates), see [`RELEASE_RUNBOOK.md`](RELEASE_RUNBOOK.md).
- [ ] `Cargo.toml` workspace `version` is bumped.
- [ ] All per-crate `crates/*/Cargo.toml` path-dependency `version = "..."`
pins match the new workspace version.
- [ ] `npm/deepseek-tui/package.json` `version` AND `deepseekBinaryVersion`
- [ ] `npm/codewhale-tui/package.json` `version` AND `deepseekBinaryVersion`
are both bumped.
- [ ] `Cargo.lock` is refreshed (`cargo update --workspace --offline`).
- [ ] `./scripts/release/check-versions.sh` reports
@@ -51,7 +51,7 @@ Run, in order, from the repo root:
## 4. npm wrapper smoke
- [ ] `cargo build --release --locked -p deepseek-tui-cli -p deepseek-tui`
- [ ] `cargo build --release --locked -p codewhale-cli -p codewhale-tui`
- [ ] `node scripts/release/npm-wrapper-smoke.js`
(Set `DEEPSEEK_TUI_KEEP_SMOKE_DIR=1` if you need to inspect the temp
install afterwards.)
@@ -82,7 +82,7 @@ Run, in order, from the repo root:
- [ ] `git push origin vX.Y.Z`
- [ ] The `release.yml` workflow has built and uploaded artifacts to the
GitHub release for this tag.
- [ ] `npm view deepseek-tui@X.Y.Z version deepseekBinaryVersion --json`
- [ ] `npm view codewhale-tui@X.Y.Z version deepseekBinaryVersion --json`
reports the new version on the npm registry.
- [ ] `crates.io` has the new version (or the `publish-crates.sh` job has
pushed it).
docs/RELEASE_RUNBOOK.md
+54 -54
View File
@@ -1,37 +1,37 @@
# DeepSeek TUI Release Runbook
# codewhale Release Runbook
This runbook is the source of truth for shipping Rust crates, GitHub release assets,
and the `deepseek-tui` npm wrapper.
and the `codewhale-tui` npm wrapper.
Current packaging note:
- `deepseek-tui` is the live runtime and TUI package shipped to users today.
- `deepseek-tui-core` is a supporting workspace crate for the extraction/parity effort, not a replacement for the shipping runtime.
- `codewhale-tui` is the live runtime and TUI package shipped to users today.
- `codewhale-tui-core` is a supporting workspace crate for the extraction/parity effort, not a replacement for the shipping runtime.
## Canonical Publish Targets
- End-user crates:
- `deepseek-tui`
- `deepseek-tui-cli`
- `codewhale-tui`
- `codewhale-cli`
- Supporting crates published from this workspace:
- `deepseek-secrets`
- `deepseek-config`
- `deepseek-protocol`
- `deepseek-state`
- `deepseek-agent`
- `deepseek-execpolicy`
- `deepseek-hooks`
- `deepseek-mcp`
- `deepseek-tools`
- `deepseek-core`
- `deepseek-app-server`
- `deepseek-tui-core`
- `deepseek-cli` on crates.io is an unrelated crate and is not part of this release flow.
- `codewhale-secrets`
- `codewhale-config`
- `codewhale-protocol`
- `codewhale-state`
- `codewhale-agent`
- `codewhale-execpolicy`
- `codewhale-hooks`
- `codewhale-mcp`
- `codewhale-tools`
- `codewhale-core`
- `codewhale-app-server`
- `codewhale-tui-core`
- `codewhale-cli` on crates.io is an unrelated crate and is not part of this release flow.
## Version Coordination
- Rust crates inherit the shared workspace version from [Cargo.toml](../Cargo.toml).
- Internal path dependency versions should match the shared workspace version; stale older pins are release blockers once the workspace version moves.
- The npm wrapper version lives in [npm/deepseek-tui/package.json](../npm/deepseek-tui/package.json).
- The npm wrapper version lives in [npm/codewhale-tui/package.json](../npm/codewhale-tui/package.json).
- `deepseekBinaryVersion` controls which GitHub release binaries the npm wrapper downloads.
- Packaging-only npm releases are allowed:
- bump the npm package version
@@ -48,13 +48,13 @@ cargo fmt --all -- --check
cargo check --workspace --all-targets --locked
cargo clippy --workspace --all-targets --all-features --locked -- -D warnings
cargo test --workspace --all-features --locked
cargo publish --dry-run --locked --allow-dirty -p deepseek-tui
cargo publish --dry-run --locked --allow-dirty -p codewhale-tui
./scripts/release/publish-crates.sh dry-run
```
`check-versions.sh` also runs in CI on every push/PR (the `versions` job in
`.github/workflows/ci.yml`), so drift between `Cargo.toml`, the per-crate
manifests, `npm/deepseek-tui/package.json`, and `Cargo.lock` is caught before
manifests, `npm/codewhale-tui/package.json`, and `Cargo.lock` is caught before
release time rather than at it.
`publish-crates.sh dry-run` performs a full `cargo publish --dry-run` for crates
@@ -65,11 +65,11 @@ new workspace version while still validating package contents before publish.
For npm wrapper verification, build the two shipped binaries and run the
cross-platform smoke harness. This packs the npm wrapper, installs it into a
clean temporary project, serves local release assets over HTTP, and checks both
the dispatcher-to-TUI path (`deepseek doctor --help`) and the direct TUI
entrypoint (`deepseek-tui --help`).
the dispatcher-to-TUI path (`codewhale doctor --help`) and the direct TUI
entrypoint (`codewhale-tui --help`).
```bash
cargo build --release --locked -p deepseek-tui-cli -p deepseek-tui
cargo build --release --locked -p codewhale-cli -p codewhale-tui
node scripts/release/npm-wrapper-smoke.js
```
@@ -81,7 +81,7 @@ directory with a full asset matrix fixture before starting the server:
```bash
DEEPSEEK_TUI_PREPARE_ALL_ASSETS=1 node scripts/release/prepare-local-release-assets.js
cd npm/deepseek-tui
cd npm/codewhale-tui
DEEPSEEK_TUI_VERSION=X.Y.Z DEEPSEEK_TUI_RELEASE_BASE_URL=http://127.0.0.1:8123/ npm run release:check
```
@@ -96,8 +96,8 @@ After publishing, prove the release is visible in both registries:
./scripts/release/check-published.sh X.Y.Z
```
Do not mark a Rust release complete until that command sees `deepseek-tui@X.Y.Z`
on npm and every `deepseek-*` crate at `X.Y.Z` on crates.io. For a rare
Do not mark a Rust release complete until that command sees `codewhale-tui@X.Y.Z`
on npm and every `codewhale-*` crate at `X.Y.Z` on crates.io. For a rare
npm packaging-only release, run with `--allow-npm-binary-mismatch` and keep the
release notes explicit that no new Rust binary version shipped.
@@ -115,20 +115,20 @@ configured.
`main` and letting `auto-tag.yml` create the tag — see the npm wrapper
release section below for the `RELEASE_TAG_PAT` requirement).
4. Publish crates in this order with `./scripts/release/publish-crates.sh publish`:
- `deepseek-secrets`
- `deepseek-config`
- `deepseek-protocol`
- `deepseek-state`
- `deepseek-agent`
- `deepseek-execpolicy`
- `deepseek-hooks`
- `deepseek-mcp`
- `deepseek-tools`
- `deepseek-core`
- `deepseek-app-server`
- `deepseek-tui-core`
- `deepseek-tui-cli`
- `deepseek-tui`
- `codewhale-secrets`
- `codewhale-config`
- `codewhale-protocol`
- `codewhale-state`
- `codewhale-agent`
- `codewhale-execpolicy`
- `codewhale-hooks`
- `codewhale-mcp`
- `codewhale-tools`
- `codewhale-core`
- `codewhale-app-server`
- `codewhale-tui-core`
- `codewhale-cli`
- `codewhale-tui`
5. Wait for each published crate version to appear on crates.io before publishing dependents.
The publish helper is idempotent for reruns: already-published crate versions are skipped.
@@ -137,16 +137,16 @@ The publish helper is idempotent for reruns: already-published crate versions ar
`.github/workflows/release.yml` builds these binaries:
- `deepseek-linux-x64`
- `deepseek-macos-x64`
- `deepseek-macos-arm64`
- `deepseek-windows-x64.exe`
- `deepseek-tui-linux-x64`
- `deepseek-tui-macos-x64`
- `deepseek-tui-macos-arm64`
- `deepseek-tui-windows-x64.exe`
- `codewhale-linux-x64`
- `codewhale-macos-x64`
- `codewhale-macos-arm64`
- `codewhale-windows-x64.exe`
- `codewhale-tui-linux-x64`
- `codewhale-tui-macos-x64`
- `codewhale-tui-macos-arm64`
- `codewhale-tui-windows-x64.exe`
The release job also uploads `deepseek-artifacts-sha256.txt`. The npm installer and
The release job also uploads `codewhale-artifacts-sha256.txt`. The npm installer and
release verification script both depend on that checksum manifest.
## npm Wrapper Release
@@ -159,14 +159,14 @@ on a workstation with `npm login` and an authenticator app.
### Steps
1. Set the npm package version in [npm/deepseek-tui/package.json](../npm/deepseek-tui/package.json) to match the workspace `Cargo.toml`. CI's version-drift guard will catch mismatches before tag.
1. Set the npm package version in [npm/codewhale-tui/package.json](../npm/codewhale-tui/package.json) to match the workspace `Cargo.toml`. CI's version-drift guard will catch mismatches before tag.
2. Set `deepseekBinaryVersion` to the GitHub release tag that should supply binaries.
3. Push the version bump to `main`. `auto-tag.yml` creates the matching `vX.Y.Z` tag, and `release.yml` builds the binary matrix and drafts the GitHub Release.
4. **Wait for the GitHub Release to finalize** with all eight signed binaries plus `deepseek-artifacts-sha256.txt`. The npm `prepublishOnly` hook (`scripts/verify-release-assets.js`) requires every asset to be present.
4. **Wait for the GitHub Release to finalize** with all eight signed binaries plus `codewhale-artifacts-sha256.txt`. The npm `prepublishOnly` hook (`scripts/verify-release-assets.js`) requires every asset to be present.
5. From a developer machine, publish the npm wrapper manually:
```bash
cd npm/deepseek-tui
cd npm/codewhale-tui
npm publish --access public
# (you will be prompted for the npm OTP from your authenticator)
```
docs/RUNTIME_API.md
+18 -18
View File
@@ -1,8 +1,8 @@
# Runtime API & Integration Contract
DeepSeek TUI exposes a local runtime API through `deepseek serve --http` and
machine-readable health via `deepseek doctor --json`. It also exposes
`deepseek serve --acp` for editor clients that speak the Agent Client Protocol
codewhale exposes a local runtime API through `codewhale serve --http` and
machine-readable health via `codewhale doctor --json`. It also exposes
`codewhale serve --acp` for editor clients that speak the Agent Client Protocol
over stdio. This document is the stable integration contract for native macOS
workbench applications (and other local supervisors) that embed the DeepSeek
engine without screen-scraping terminal output.
@@ -12,19 +12,19 @@ engine without screen-scraping terminal output.
```
macOS workbench (or any local supervisor)
│
├─ deepseek doctor --json → machine-readable health & capability
├─ deepseek serve --http → HTTP/SSE runtime API
├─ deepseek serve --acp → ACP stdio agent for editors such as Zed
├─ deepseek serve --mcp → MCP stdio server
└─ deepseek [args] → interactive TUI session
├─ codewhale doctor --json → machine-readable health & capability
├─ codewhale serve --http → HTTP/SSE runtime API
├─ codewhale serve --acp → ACP stdio agent for editors such as Zed
├─ codewhale serve --mcp → MCP stdio server
└─ codewhale [args] → interactive TUI session
```
The engine runs as a local-only process. All APIs bind to `localhost` by
default. No hosted relay, no provider-token custody, no secret leakage.
## ACP stdio adapter: `deepseek serve --acp`
## ACP stdio adapter: `codewhale serve --acp`
`deepseek serve --acp` speaks JSON-RPC 2.0 over newline-delimited stdio for
`codewhale serve --acp` speaks JSON-RPC 2.0 over newline-delimited stdio for
ACP-compatible editor clients. The initial adapter implements the ACP baseline:
- `initialize`
@@ -38,16 +38,16 @@ followed by a `session/prompt` response with `stopReason: "end_turn"`.
The adapter is intentionally conservative: it does not yet expose shell tools,
file-write tools, checkpoint replay, or session loading through ACP. Use
`deepseek serve --http` for the full local runtime API and `deepseek serve --mcp`
`codewhale serve --http` for the full local runtime API and `codewhale serve --mcp`
when another client needs DeepSeek's tools as MCP tools.
## Capability endpoint: `deepseek doctor --json`
## Capability endpoint: `codewhale doctor --json`
Returns a JSON object describing the current installation's readiness state.
Suitable for health-check polling from a macOS workbench.
```bash
deepseek doctor --json
codewhale doctor --json
```
### Response schema (key fields)
@@ -88,7 +88,7 @@ deepseek doctor --json
"version": "0.8.9",
"config_path": "/Users/you/.deepseek/config.toml",
"config_present": true,
"workspace": "/Users/you/projects/deepseek-tui",
"workspace": "/Users/you/projects/codewhale-tui",
"api_key": {
"source": "env"
},
@@ -113,10 +113,10 @@ deepseek doctor --json
}
```
## HTTP/SSE runtime API: `deepseek serve --http`
## HTTP/SSE runtime API: `codewhale serve --http`
```bash
deepseek serve --http [--host 127.0.0.1] [--port 7878] [--workers 2] [--auth-token TOKEN]
codewhale serve --http [--host 127.0.0.1] [--port 7878] [--workers 2] [--auth-token TOKEN]
```
Defaults: host `127.0.0.1`, port `7878`, 2 workers (clamped 1–8).
@@ -333,7 +333,7 @@ The runtime API ships with a built-in dev-origin allow-list:
`http://127.0.0.1:1420`, `tauri://localhost`. To add additional origins (e.g.
when developing a UI on Vite's default `:5173`), use any of:
- CLI flag (repeatable): `deepseek serve --http --cors-origin http://localhost:5173`
- CLI flag (repeatable): `codewhale serve --http --cors-origin http://localhost:5173`
- Env var (comma-separated): `DEEPSEEK_CORS_ORIGINS="http://localhost:5173,http://localhost:8080"`
- Config (`~/.deepseek/config.toml`):
```toml
@@ -366,7 +366,7 @@ model is preserved. Added in v0.8.10 (#561).
Contract snapshots live in `crates/protocol/tests/`. Run:
```bash
cargo test -p deepseek-protocol --test parity_protocol --locked
cargo test -p codewhale-protocol --test parity_protocol --locked
```
This validates that the app-server's event schema hasn't drifted from the
docs/TENCENT_CLOUD_REMOTE_FIRST.md
+14 -14
View File
@@ -1,10 +1,10 @@
# Tencent Cloud Remote-First Quickstart
This is the opinionated Tencent-native teaching path for DeepSeek TUI users
This is the opinionated Tencent-native teaching path for codewhale users
who want an always-on agent workspace, a phone control surface, and a stack
that works well from mainland China.
It complements the local install path. If you only want to use `deepseek` on a
It complements the local install path. If you only want to use `codewhale` on a
laptop, start with the README quickstart. If you want "DS-TUI as a remote
workbench I can control from my phone", start here.
@@ -15,10 +15,10 @@ GitHub main/tags
-> CNB mirror: cnb.cool/deepseek-tui.com/DeepSeek-TUI
-> optional CNB build/deploy pipeline
-> Tencent Lighthouse HK
/opt/whalebro/deepseek-tui
/opt/whalebro/codewhale-tui
/opt/whalebro/worktrees
deepseek-runtime.service on 127.0.0.1:7878
deepseek-feishu-bridge.service
codewhale-runtime.service on 127.0.0.1:7878
codewhale-feishu-bridge.service
-> Feishu/Lark phone DM
EdgeOne is optional:
@@ -32,7 +32,7 @@ EdgeOne is optional:
slow. Optional CNB deploy templates live under
`deploy/tencent-lighthouse/cnb/`.
- **Lighthouse** is the private always-on host. It owns `/opt/whalebro`,
systemd, Rust/Node installs, and the `deepseek serve --http` runtime.
systemd, Rust/Node installs, and the `codewhale serve --http` runtime.
- **Feishu/Lark** is the first phone UI. The bridge uses long-connection mode,
so the first setup does not need a public webhook URL.
- **EdgeOne** is the public edge only when you intentionally expose a web
@@ -57,14 +57,14 @@ EdgeOne is optional:
```bash
export DEEPSEEK_BRANCH=main
git clone --branch "$DEEPSEEK_BRANCH" "$DEEPSEEK_REPO_URL" /tmp/deepseek-tui
cd /tmp/deepseek-tui
git clone --branch "$DEEPSEEK_BRANCH" "$DEEPSEEK_REPO_URL" /tmp/codewhale-tui
cd /tmp/codewhale-tui
sudo DEEPSEEK_REPO_URL="$DEEPSEEK_REPO_URL" \
DEEPSEEK_REPO_BRANCH="$DEEPSEEK_BRANCH" \
bash scripts/tencent-lighthouse/bootstrap-ubuntu.sh
```
4. Install Rust for the `deepseek` user, build both binaries, and install the
4. Install Rust for the `codewhale` user, build both binaries, and install the
systemd units using `docs/TENCENT_LIGHTHOUSE_HK.md`.
5. Configure a Feishu/Lark self-built app, fill
`/etc/deepseek/feishu-bridge.env`, run the validator, then run the VPS
@@ -85,7 +85,7 @@ The intended deploy button should:
1. Run bridge validation/tests and lightweight release-version checks.
2. SSH to Lighthouse with a deploy key stored as a CNB secret.
3. Update `/opt/whalebro/deepseek-tui`.
3. Update `/opt/whalebro/codewhale-tui`.
4. Rebuild/install both binaries.
5. Reinstall/restart systemd services.
6. Run `scripts/tencent-lighthouse/doctor.sh`.
@@ -105,7 +105,7 @@ you want a public domain in front of a deliberate HTTP service:
Keep these rules:
- `deepseek serve --http` stays bound to `127.0.0.1`.
- `codewhale serve --http` stays bound to `127.0.0.1`.
- `/v1/*` runtime endpoints are never public.
- `DEEPSEEK_RUNTIME_TOKEN` never leaves the server env files.
- Feishu/Lark group control stays off until a specific group allowlist is set.
@@ -114,13 +114,13 @@ Keep these rules:
## Teaching Order
Use this sequence when explaining DeepSeek TUI to a new remote-first user:
Use this sequence when explaining codewhale to a new remote-first user:
1. **Local mental model:** `deepseek` is the dispatcher, `deepseek-tui` is the
1. **Local mental model:** `codewhale` is the dispatcher, `codewhale-tui` is the
companion runtime, and both binaries matter.
2. **Agent safety:** Plan/Agent/YOLO are separate from approval mode and
sandboxing.
3. **Remote runtime:** `deepseek serve --http` is a localhost runtime API, not
3. **Remote runtime:** `codewhale serve --http` is a localhost runtime API, not
a public web app.
4. **Phone bridge:** Feishu/Lark messages become runtime requests through an
allowlisted bridge.
docs/TENCENT_LIGHTHOUSE_HK.md
+24 -24
View File
@@ -1,7 +1,7 @@
# Tencent Lighthouse Hong Kong Phone Setup
This runbook sets up a Tencent Cloud Lighthouse instance in Hong Kong as an
always-on DeepSeek TUI host controlled from Feishu/Lark on a phone.
always-on codewhale host controlled from Feishu/Lark on a phone.
If you are teaching this as the Tencent-native default path, start with
[docs/TENCENT_CLOUD_REMOTE_FIRST.md](TENCENT_CLOUD_REMOTE_FIRST.md). This file
@@ -11,14 +11,14 @@ is the implementation runbook for the Lighthouse host itself.
```text
CNB mirror or GitHub branch
-> /opt/whalebro/deepseek-tui
-> /opt/whalebro/codewhale-tui
Feishu/Lark mobile app
-> Feishu/Lark long-connection bot
-> deepseek-feishu-bridge systemd service
-> http://127.0.0.1:7878 deepseek serve --http
-> codewhale-feishu-bridge systemd service
-> http://127.0.0.1:7878 codewhale serve --http
-> /opt/whalebro
-> deepseek-tui/
-> codewhale-tui/
Optional public edge:
EdgeOne -> Caddy/Nginx public site on Lighthouse
@@ -31,11 +31,11 @@ HTTP service, not the runtime API.
## Remote Whalebro Workspace
Use `/opt/whalebro` as the VPS workspace root. The first-class checkout is
`/opt/whalebro/deepseek-tui`.
`/opt/whalebro/codewhale-tui`.
Create these paths first:
- `/opt/whalebro/deepseek-tui`
- `/opt/whalebro/codewhale-tui`
- `/opt/whalebro/worktrees`
Linux is enough for Rust, Node, and service work. Mac-only release work such
@@ -88,8 +88,8 @@ sudo apt-get update
sudo apt-get install -y git
export DEEPSEEK_BRANCH=main
export DEEPSEEK_REPO_URL=https://cnb.cool/deepseek-tui.com/DeepSeek-TUI.git
git clone --branch "$DEEPSEEK_BRANCH" "$DEEPSEEK_REPO_URL" /tmp/deepseek-tui
cd /tmp/deepseek-tui
git clone --branch "$DEEPSEEK_BRANCH" "$DEEPSEEK_REPO_URL" /tmp/codewhale-tui
cd /tmp/codewhale-tui
sudo DEEPSEEK_REPO_URL="$DEEPSEEK_REPO_URL" \
DEEPSEEK_REPO_BRANCH="$DEEPSEEK_BRANCH" \
bash scripts/tencent-lighthouse/bootstrap-ubuntu.sh
@@ -120,16 +120,16 @@ If this deployment setup has not been pushed to Git yet, either push the branch
first or copy this checkout to the VPS before running these commands. A fresh
VPS clone cannot see uncommitted local files.
Install Rust 1.88+ for the `deepseek` user, then build both shipped binaries:
Install Rust 1.88+ for the `codewhale` user, then build both shipped binaries:
```bash
sudo -iu deepseek
sudo -iu codewhale
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs -o /tmp/rustup-init.sh
sed -n '1,120p' /tmp/rustup-init.sh
sh /tmp/rustup-init.sh -y --profile minimal
. "$HOME/.cargo/env"
rustup default stable
cd /opt/whalebro/deepseek-tui
cd /opt/whalebro/codewhale-tui
cargo install --path crates/cli --locked --force
cargo install --path crates/tui --locked --force
exit
@@ -138,14 +138,14 @@ exit
Copy and install the bridge/service files:
```bash
cd /opt/whalebro/deepseek-tui
cd /opt/whalebro/codewhale-tui
sudo bash scripts/tencent-lighthouse/install-services.sh
```
After editing both env files, validate the bridge/runtime pairing:
```bash
sudo -u deepseek node /opt/deepseek/bridge/scripts/validate-config.mjs \
sudo -u codewhale node /opt/deepseek/bridge/scripts/validate-config.mjs \
--env /etc/deepseek/feishu-bridge.env \
--runtime-env /etc/deepseek/runtime.env \
--workspace-root /opt/whalebro \
@@ -185,25 +185,25 @@ For first pairing, either:
## Start Services
```bash
sudo systemctl start deepseek-runtime
sudo systemctl status deepseek-runtime --no-pager
sudo systemctl start codewhale-runtime
sudo systemctl status codewhale-runtime --no-pager
curl -s http://127.0.0.1:7878/health
sudo systemctl start deepseek-feishu-bridge
sudo journalctl -u deepseek-feishu-bridge -f
sudo systemctl start codewhale-feishu-bridge
sudo journalctl -u codewhale-feishu-bridge -f
```
Run the Lighthouse doctor after both services are configured:
```bash
cd /opt/whalebro/deepseek-tui
cd /opt/whalebro/codewhale-tui
sudo bash scripts/tencent-lighthouse/doctor.sh
```
Enable on boot is done by `install-services.sh`; if needed:
```bash
sudo systemctl enable deepseek-runtime deepseek-feishu-bridge
sudo systemctl enable codewhale-runtime codewhale-feishu-bridge
```
## Phone Commands
@@ -280,14 +280,14 @@ From a phone DM to the bot:
5. Trigger a tool approval and verify both `/allow <approval_id>` and
`/deny <approval_id>` paths.
6. Restart both services and re-run `/status`.
7. Reboot the instance, then confirm `systemctl status deepseek-runtime` and
`systemctl status deepseek-feishu-bridge` return to active.
7. Reboot the instance, then confirm `systemctl status codewhale-runtime` and
`systemctl status codewhale-feishu-bridge` return to active.
## Operational Notes
- Bind `deepseek serve --http` to `127.0.0.1`.
- Bind `codewhale serve --http` to `127.0.0.1`.
- Keep the Lighthouse firewall focused on SSH for this setup.
- Use SSH key auth.
- Use `tmux` for emergency terminal work from Blink/Termius.
- Keep `/opt/whalebro/deepseek-tui` on a personal branch while working from the
- Keep `/opt/whalebro/codewhale-tui` on a personal branch while working from the
phone.
docs/TOOL_SURFACE.md
+2 -2
View File
@@ -258,8 +258,8 @@ while the registry contract stays stable.
Version smoke:
```bash
deepseek --version
deepseek-tui --version
codewhale --version
codewhale-tui --version
```
Tool-surface smoke:
docs/capacity_controller.md
+1 -1
View File
@@ -1,6 +1,6 @@
# Capacity Controller
`deepseek-tui` includes an opt-in capacity-aware context controller. In the
`codewhale-tui` includes an opt-in capacity-aware context controller. In the
default V4 path it is disabled, because its active interventions can rewrite
the live prompt and break prefix-cache affinity. Treat it as telemetry or an
experimental guardrail unless `capacity.enabled = true` is set explicitly.