agentsw 0.3.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 luocheng
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,227 @@
+<div align="center">
+
+# agentsw
+
+**One provider, every coding agent — from the terminal.**
+
+A CLI that keeps your OpenAI- and Anthropic-protocol providers — base URL, key, model list,
+metadata — in one place and writes the active one into nine agents' own config files,
+without touching the settings you hand-tuned.
+
+[![npm version](https://img.shields.io/npm/v/agentsw?logo=npm&logoColor=white)](https://www.npmjs.com/package/agentsw)
+[![CI](https://img.shields.io/github/actions/workflow/status/tchivs/agentsw/ci.yml?branch=main&logo=github&logoColor=white&label=CI)](https://github.com/tchivs/agentsw/actions/workflows/ci.yml)
+[![node version](https://img.shields.io/node/v/agentsw?logo=nodedotjs&logoColor=white)](#install)
+[![license](https://img.shields.io/npm/l/agentsw?color=blue)](./LICENSE)
+
+[Install](#install) · [Quick start](#quick-start) · [Supported apps](#supported-apps) · [Commands](#commands) · [简体中文](./README.zh-CN.md)
+
+</div>
+
+```console
+$ asw use vfing -m glm-5.3-flash
+switching to vfing (openai) · default model glm-5.3-flash
+
+skip claude    Claude Code does not support openai-protocol providers
+ok   codex     ~/.codex/config.toml, ~/.codex/auth.json
+               backup: ~/.config/agentsw/backups/config.toml.2026-08-31T15-58-24-678Z
+ok   omp       ~/.omp/agent/models.yml
+               select in omp with: omp --model vfing/glm-5.3-flash
+skip pi        pi not detected (pass --apps pi to force)
+ok   opencode  ~/.config/opencode/opencode.json
+ok   dsh       ~/.dsh/settings.yaml, ~/.dsh/.credentials.yaml
+               select in dsh with the model picker, or run: dsh web
+```
+
+`agentsw` and the short `asw` are the same binary.
+
+## Why a CLI
+
+- **It runs where a desktop app cannot.** Headless servers, containers, devcontainers, CI.
+  `npx agentsw use myproxy -a codex,omp` is one line in a provisioning script; nine GUI
+  clicks are not.
+- **`--dry-run` prints diffs.** Review what a switch will write before it writes it, or pipe
+  it into a code review. Every file is backed up first regardless.
+- **Non-destructive by design.** Only the fields agentsw owns are rewritten. Provider-level
+  keys it does not model — `authHeader`, `headers`, `compat`, `discovery` — per-model extras
+  like `thinkingLevelMap`, and YAML comments all survive a re-sync.
+- **Model metadata, not just an id list.** Reseller `/v1/models` endpoints return bare ids.
+  agentsw enriches them from [models.dev](https://models.dev) and writes what each app
+  actually reads: `thinkingLevelMap` for pi/prime, `limit`/`attachment` for opencode,
+  `contextWindow`/`maxTokens` for omp, `reasoningEfforts`/`input` for dsh.
+- **Adopts what you already have** — including [cc-switch](https://github.com/farion1231/cc-switch).
+  `import` scans every installed agent *and* cc-switch's own store, merges duplicates
+  (same protocol + base URL), and never writes back to either.
+- **Wire-aware.** `/v1/chat/completions` and `/v1/responses` are different endpoints;
+  agentsw tracks which one a provider speaks and never downgrades a working one.
+
+Want a GUI, MCP/Skills sync, a usage dashboard or a local failover proxy? Use
+[cc-switch](https://github.com/farion1231/cc-switch) — it is excellent at that, and agentsw
+imports from it rather than competing with it.
+
+## Install
+
+```bash
+npx agentsw               # zero install: opens the interactive menu
+npm install -g agentsw    # requires Node >= 22.12
+```
+
+## Quick start
+
+```bash
+agentsw                   # no args: interactive menu (add, import, use, status, apps)
+```
+
+The first run on an empty store offers to import the providers already configured in your
+agents, and asks for English or 简体中文. Prefer flags? The same flow, unattended:
+
+```bash
+# add a provider and let it discover its own model list
+asw add -y --id vfing --protocol openai --openai-api responses \
+  --base-url https://api.example.com/v1 --api-key sk-... --discover
+
+asw use vfing             # point every detected agent at it
+asw use vfing --dry-run   # ...or just show the diffs
+asw status                # what each agent points at right now
+```
+
+```console
+$ asw status
+config: ~/.config/agentsw/config.json
+active provider: vfing
+
+APP        FOUND  PROTOCOLS         CURRENT                               CONFIG
+---------  -----  ----------------  ------------------------------------  ---------------------------
+claude     yes    anthropic         -                                     ~/.claude/settings.json
+codex      yes    openai            vfing · glm-5.3-flash                 ~/.codex/config.toml
+omp        yes    openai+anthropic  providers: sub, sub-anthropic, vfing  ~/.omp/agent/models.yml
+pi         no     openai+anthropic  -                                     ~/.pi/agent/models.json
+opencode   yes    openai+anthropic  vfing/glm-5.3-flash                   ~/.config/opencode/opencode.json
+dsh        yes    openai+anthropic  vfing · glm-5.3-flash                 ~/.dsh/settings.yaml
+```
+
+## Supported apps
+
+| App | Config written | Protocols |
+|---|---|---|
+| [Claude Code](https://www.npmjs.com/package/@anthropic-ai/claude-code) | `~/.claude/settings.json` env block | anthropic |
+| [Codex CLI](https://www.npmjs.com/package/@openai/codex) | `~/.codex/config.toml` + `auth.json` (Responses API only) | openai |
+| [Oh My Pi](https://omp.sh) (omp) | `~/.omp/agent/models.yml` (comments preserved) | both |
+| [pi](https://www.npmjs.com/package/@earendil-works/pi-coding-agent) | `~/.pi/agent/models.json` + `settings.json` | both |
+| [prime-agent](https://github.com/PrimeIntellect-ai/prime-agent) | `~/.prime/agent/models.json` + `settings.json` | both |
+| [opencode](https://opencode.ai) | `~/.config/opencode/opencode.json` | both |
+| [Hermes](https://pypi.org/project/hermes-agent/) | `~/.hermes/config.yaml` + `.env` (comments preserved) | both |
+| WorkBuddy | `~/.workbuddy/models.json` + `settings.json` | openai |
+| [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) (dsh) | `~/.dsh/settings.yaml` + `.credentials.yaml` | both |
+
+An agent that is not installed is skipped, not guessed at. `--apps codex,omp` narrows a run;
+`--apps` with an uninstalled agent forces it.
+
+## Import what you already configured
+
+`import` reads every agent's config **and cc-switch's SQLite store** (read-only), resolves
+inline and env-backed keys, and merges candidates that share a protocol + base URL — model
+lists and source apps are unioned. Different protocols on the same host stay separate.
+
+```console
+$ asw import --all
+ID             PROTOCOL   BASE URL                                MODELS  FROM       KEY
+-------------  ---------  --------------------------------------  ------  ---------  ---
+any            anthropic  https://a-ocnfniawgw.cn-shanghai...     1       cc-switch  yes
+sub            openai     https://new.vfing.de/v1                 18      omp        yes
+sub-anthropic  anthropic  https://new.vfing.de                    18      omp        yes
+zhipu-glm      anthropic  https://open.bigmodel.cn/api/anthropic  1       cc-switch  yes
+imported sub · openai · https://new.vfing.de/v1 · 18 models [from omp]
+imported zhipu-glm · anthropic · https://open.bigmodel.cn/api/anthropic · 1 models [from cc-switch]
+```
+
+Without `--all` the list is a multi-select preview; nothing is written until you pick.
+
+## Model discovery filters
+
+`--discover` lists model ids from the provider's `/v1/models`, then enriches each id from
+models.dev. Reseller lists are noisy, so filters are persisted per provider and re-applied on
+every `discover`:
+
+```bash
+# snapshot duplicates (gpt-5.2-latest, glm-4.7-250414, ...) are DROPPED by default
+# whenever the bare id is also listed; snapshot-only models are kept as-is
+asw add -y --discover --exclude "*embedding*,*video*" --id myproxy ...
+asw add -y --discover --no-dedup ...            # keep duplicates
+asw discover myproxy --include "gpt-*,glm-*"    # update the persisted filter
+asw discover myproxy --no-filter                # clear it
+```
+
+Explicit `--models` entries and the default model are never filtered out.
+
+## Commands
+
+| Command | What it does |
+|---|---|
+| `agentsw` | interactive menu (no arguments) |
+| `add` | add or update a provider; `--discover` fills the model list |
+| `import [--all]` | adopt providers from your agents' configs and cc-switch |
+| `list` / `status` | configured providers / what each agent points at |
+| `use <id>` | switch every detected agent; `-a codex,omp`, `-m <model>`, `--dry-run` |
+| `sync` | re-apply the active provider (after an agent update, say) |
+| `discover <id> [--sync]` | refresh the model list + metadata from `/v1/models` |
+| `models [query]` | search the models.dev catalog |
+| `refresh` | re-fetch metadata for every configured provider |
+| `prune <id>` / `remove <id> [--prune]` | remove from app configs / from the store |
+| `apps` / `install <app>` / `upgrade` | agent version manager |
+
+```bash
+asw use myproxy -a codex,omp -m glm-5.2
+asw discover myproxy --sync
+asw remove myproxy --prune
+```
+
+## How your configs are treated
+
+- Every modified file is backed up first to `~/.config/agentsw/backups/`.
+- The provider store lives at `~/.config/agentsw/config.json` (mode 0600).
+- A sync overwrites only the fields agentsw owns. Unmodeled provider-level keys and per-model
+  extras survive; an owned field that stops applying is cleared rather than left stale, and a
+  per-model `api`/`baseUrl` override that contradicts the route is dropped (it would silently
+  win over the entry) and reported.
+- YAML configs (omp, Hermes, dsh) keep their comments. `~/.codex/config.toml` does not —
+  TOML round-trip drops them, which is why the backup happens first.
+- OpenAI endpoints come in two wires: `/v1/chat/completions` and `/v1/responses`. Imported
+  providers keep whatever the config declared — on the provider or on its models — the
+  interactive `add` asks, and `--openai-api responses` sets it by hand. A sync never
+  downgrades an existing responses entry.
+- Codex only speaks the Responses API; a chat-completions-only endpoint cannot be used with
+  it, and the sync output says so.
+- DeepSeek Harness keeps secrets out of `settings.yaml`: the route carries
+  `apiKeyEnv: AGENTSW_<ID>_API_KEY`, the key goes into `$DSH_HOME/.credentials.yaml`
+  (`refs:`, mode 0600). `$DSH_HOME` defaults to `~/.dsh`.
+- cc-switch's database is opened read-only and never written to.
+- The apps manager (install/upgrade) currently targets macOS/Linux.
+
+## Language
+
+The first interactive launch asks for English or 简体中文 and saves the choice; the menu can
+change it later. A one-run override wins over the saved choice, and with neither the system
+locale decides.
+
+```bash
+asw --lang zh-CN
+AGENTSW_LANG=en asw
+```
+
+## Development
+
+```bash
+npm install
+npm run build       # tsc -> dist/
+npm test            # node:test suite (filter semantics, adapter roundtrips)
+npm run dev -- ...  # run from source via tsx
+```
+
+Each agent is one adapter in [`src/targets/`](./src/targets) implementing
+`apply` / `prune` / `current` / `candidates`; read-only import sources live in
+[`src/sources/`](./src/sources). Adding an app is one file plus a line in
+[`src/targets/index.ts`](./src/targets/index.ts).
+
+## License
+
+MIT

package/README.zh-CN.md

@@ -0,0 +1,221 @@
+<div align="center">
+
+# agentsw
+
+**一个供应商，管遍所有编码智能体——在终端里。**
+
+一个命令行工具：把 OpenAI 协议 / Anthropic 协议的供应商——base URL、密钥、模型列表、
+元数据——集中管理，再把当前选中的那个写进九个智能体各自的配置文件，
+且不动你手工调过的设置。
+
+[![npm version](https://img.shields.io/npm/v/agentsw?logo=npm&logoColor=white)](https://www.npmjs.com/package/agentsw)
+[![CI](https://img.shields.io/github/actions/workflow/status/tchivs/agentsw/ci.yml?branch=main&logo=github&logoColor=white&label=CI)](https://github.com/tchivs/agentsw/actions/workflows/ci.yml)
+[![node version](https://img.shields.io/node/v/agentsw?logo=nodedotjs&logoColor=white)](#安装)
+[![license](https://img.shields.io/npm/l/agentsw?color=blue)](./LICENSE)
+
+[安装](#安装) · [快速开始](#快速开始) · [支持的应用](#支持的应用) · [命令](#命令) · [English](./README.md)
+
+</div>
+
+```console
+$ asw use vfing -m glm-5.3-flash
+switching to vfing (openai) · default model glm-5.3-flash
+
+skip claude    Claude Code does not support openai-protocol providers
+ok   codex     ~/.codex/config.toml, ~/.codex/auth.json
+               backup: ~/.config/agentsw/backups/config.toml.2026-08-31T15-58-24-678Z
+ok   omp       ~/.omp/agent/models.yml
+               select in omp with: omp --model vfing/glm-5.3-flash
+skip pi        pi not detected (pass --apps pi to force)
+ok   opencode  ~/.config/opencode/opencode.json
+ok   dsh       ~/.dsh/settings.yaml, ~/.dsh/.credentials.yaml
+               select in dsh with the model picker, or run: dsh web
+```
+
+`agentsw` 和短名 `asw` 是同一个可执行文件。
+
+## 为什么是 CLI
+
+- **它能去桌面应用去不了的地方。** 无头服务器、容器、devcontainer、CI。
+  `npx agentsw use myproxy -a codex,omp` 在开机脚本里是一行；GUI 点九次不是。
+- **`--dry-run` 直接打 diff。** 写盘前先看清楚要改什么，或者把 diff 丢进 code review。
+  无论如何每个文件都会先备份。
+- **默认不破坏现有配置。** 只覆盖 agentsw 自己管理的字段：它不建模的供应商级键
+  （`authHeader`、`headers`、`compat`、`discovery`）、`thinkingLevelMap` 之类的模型级字段、
+  以及 YAML 注释，都会在重新同步后原样保留。
+- **写的是元数据，不只是模型 id。** 转售商 `/v1/models` 只返回裸 id；agentsw 从
+  [models.dev](https://models.dev) 补全，并按各应用真正读取的字段下发：pi/prime 的
+  `thinkingLevelMap`、opencode 的 `limit`/`attachment`、omp 的 `contextWindow`/`maxTokens`、
+  dsh 的 `reasoningEfforts`/`input`。
+- **接管你已有的配置——包括 [cc-switch](https://github.com/farion1231/cc-switch)。**
+  `import` 会扫描所有已安装的智能体**以及 cc-switch 自己的存储**，合并重复项
+  （相同协议 + base URL），并且对两者都只读不写。
+- **区分接口形态。** `/v1/chat/completions` 与 `/v1/responses` 是两个不同端点；
+  agentsw 记录供应商说的是哪一种，并且绝不把可用的 responses 降级。
+
+想要 GUI、MCP/Skills 同步、用量仪表盘或本地故障转移代理？用
+[cc-switch](https://github.com/farion1231/cc-switch)——那些它做得很好；agentsw 从它导入，
+而不是跟它抢。
+
+## 安装
+
+```bash
+npx agentsw               # 免安装，直接打开交互菜单
+npm install -g agentsw    # 需要 Node >= 22.12
+```
+
+## 快速开始
+
+```bash
+agentsw                   # 不带参数：交互菜单（添加、导入、切换、状态、应用管理）
+```
+
+存储为空时，首次运行会询问是否导入各智能体里已配置的供应商，并让你选择
+English 或简体中文。想用参数跑？同样的流程，全自动：
+
+```bash
+# 添加供应商，并让它自己发现模型列表
+asw add -y --id vfing --protocol openai --openai-api responses \
+  --base-url https://api.example.com/v1 --api-key sk-... --discover
+
+asw use vfing             # 把所有检测到的智能体指向它
+asw use vfing --dry-run   # ……或者只看 diff
+asw status                # 各智能体当前指向什么
+```
+
+```console
+$ asw status
+config: ~/.config/agentsw/config.json
+active provider: vfing
+
+APP        FOUND  PROTOCOLS         CURRENT                               CONFIG
+---------  -----  ----------------  ------------------------------------  ---------------------------
+claude     yes    anthropic         -                                     ~/.claude/settings.json
+codex      yes    openai            vfing · glm-5.3-flash                 ~/.codex/config.toml
+omp        yes    openai+anthropic  providers: sub, sub-anthropic, vfing  ~/.omp/agent/models.yml
+pi         no     openai+anthropic  -                                     ~/.pi/agent/models.json
+opencode   yes    openai+anthropic  vfing/glm-5.3-flash                   ~/.config/opencode/opencode.json
+dsh        yes    openai+anthropic  vfing · glm-5.3-flash                 ~/.dsh/settings.yaml
+```
+
+## 支持的应用
+
+| 应用 | 写入的配置 | 协议 |
+|---|---|---|
+| [Claude Code](https://www.npmjs.com/package/@anthropic-ai/claude-code) | `~/.claude/settings.json` env 块 | anthropic |
+| [Codex CLI](https://www.npmjs.com/package/@openai/codex) | `~/.codex/config.toml` + `auth.json`（仅 Responses API） | openai |
+| [Oh My Pi](https://omp.sh) (omp) | `~/.omp/agent/models.yml`（保留注释） | 双协议 |
+| [pi](https://www.npmjs.com/package/@earendil-works/pi-coding-agent) | `~/.pi/agent/models.json` + `settings.json` | 双协议 |
+| [prime-agent](https://github.com/PrimeIntellect-ai/prime-agent) | `~/.prime/agent/models.json` + `settings.json` | 双协议 |
+| [opencode](https://opencode.ai) | `~/.config/opencode/opencode.json` | 双协议 |
+| [Hermes](https://pypi.org/project/hermes-agent/) | `~/.hermes/config.yaml` + `.env`（保留注释） | 双协议 |
+| WorkBuddy | `~/.workbuddy/models.json` + `settings.json` | openai |
+| [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) (dsh) | `~/.dsh/settings.yaml` + `.credentials.yaml` | 双协议 |
+
+未安装的应用会被跳过，而不是瞎猜。`--apps codex,omp` 只跑指定应用；
+对未检测到的应用显式传 `--apps` 可强制写入。
+
+## 导入你已有的配置
+
+`import` 读取每个智能体的配置**以及 cc-switch 的 SQLite 存储**（只读），解析内联或环境变量
+形式的密钥，并把「协议 + base URL」相同的候选合并——模型列表与来源应用取并集。
+同一域名下的不同协议不会误合并。
+
+```console
+$ asw import --all
+ID             PROTOCOL   BASE URL                                MODELS  FROM       KEY
+-------------  ---------  --------------------------------------  ------  ---------  ---
+any            anthropic  https://a-ocnfniawgw.cn-shanghai...     1       cc-switch  yes
+sub            openai     https://new.vfing.de/v1                 18      omp        yes
+sub-anthropic  anthropic  https://new.vfing.de                    18      omp        yes
+zhipu-glm      anthropic  https://open.bigmodel.cn/api/anthropic  1       cc-switch  yes
+imported sub · openai · https://new.vfing.de/v1 · 18 models [from omp]
+imported zhipu-glm · anthropic · https://open.bigmodel.cn/api/anthropic · 1 models [from cc-switch]
+```
+
+不加 `--all` 时先展示多选预览，选定之前不写任何文件。
+
+## 模型发现过滤
+
+`--discover` 从供应商 `/v1/models` 拉取模型 id，再从 models.dev 补全元数据。
+转售商列表噪音大，过滤规则持久化在供应商上，每次 `discover` 自动复用：
+
+```bash
+# 快照重复项（gpt-5.2-latest、glm-4.7-250414 等）默认丢弃——
+# 仅当裸名也在列表中才算重复；只有快照没有裸名的模型原样保留
+asw add -y --discover --exclude "*embedding*,*video*" --id myproxy ...
+asw add -y --discover --no-dedup ...            # 保留重复项
+asw discover myproxy --include "gpt-*,glm-*"    # 更新持久化过滤规则
+asw discover myproxy --no-filter                # 清除规则
+```
+
+手动 `--models` 指定的 id 和默认模型永不被过滤。
+
+## 命令
+
+| 命令 | 作用 |
+|---|---|
+| `agentsw` | 交互菜单（不带参数） |
+| `add` | 添加或更新供应商；`--discover` 自动填模型列表 |
+| `import [--all]` | 接管各智能体配置与 cc-switch 里已有的供应商 |
+| `list` / `status` | 已配置的供应商 / 各智能体当前指向 |
+| `use <id>` | 切换所有检测到的智能体；`-a codex,omp`、`-m <model>`、`--dry-run` |
+| `sync` | 重新应用当前供应商（比如某个智能体升级之后） |
+| `discover <id> [--sync]` | 从 `/v1/models` 刷新模型列表与元数据 |
+| `models [query]` | 搜索 models.dev 目录 |
+| `refresh` | 重新拉取所有供应商的元数据 |
+| `prune <id>` / `remove <id> [--prune]` | 从各应用配置中清除 / 从存储中删除 |
+| `apps` / `install <app>` / `upgrade` | 智能体版本管理 |
+
+```bash
+asw use myproxy -a codex,omp -m glm-5.2
+asw discover myproxy --sync
+asw remove myproxy --prune
+```
+
+## 它如何对待你的配置
+
+- 每个被修改的文件先备份到 `~/.config/agentsw/backups/`。
+- 供应商存储位于 `~/.config/agentsw/config.json`（权限 0600）。
+- 同步只覆盖 agentsw 自己管理的字段：未建模的供应商级键与模型级扩展字段保留；
+  自己管理但本次不再写出的字段会被清除而非留成陈旧值；与新路由矛盾的模型级
+  `api`/`baseUrl` 覆盖会被删除（否则它会静默盖过供应商条目）并在输出中点名。
+- YAML 配置（omp、Hermes、dsh）保留注释；`~/.codex/config.toml` 不行——TOML 往返会丢注释，
+  所以才先备份。
+- OpenAI 端点有两种接口形态：`/v1/chat/completions` 与 `/v1/responses`。导入时沿用配置里
+  声明的形态（供应商级或模型级），交互式 `add` 会询问，也可用 `--openai-api responses`
+  手动指定；同步绝不把已有的 responses 端点降级。
+- Codex 只支持 Responses API；仅提供 chat-completions 的端点无法用于 Codex，同步输出会提示。
+- DeepSeek Harness 不把密钥写进 `settings.yaml`：路由里只保存
+  `apiKeyEnv: AGENTSW_<ID>_API_KEY` 引用，密钥写入 `$DSH_HOME/.credentials.yaml`
+  的 `refs:`（权限 0600）。`$DSH_HOME` 默认为 `~/.dsh`。
+- cc-switch 的数据库以只读方式打开，永不写入。
+- 安装管理器（install/upgrade）目前仅支持 macOS/Linux。
+
+## 语言
+
+首次打开交互菜单会选择 English 或简体中文并持久化，之后可在菜单中切换。
+单次覆盖优先于已保存设置；两者都没有时按系统语言判断。
+
+```bash
+asw --lang zh-CN
+AGENTSW_LANG=en asw
+```
+
+## 开发
+
+```bash
+npm install
+npm run build       # tsc -> dist/
+npm test            # node:test 测试（过滤语义、适配器往返）
+npm run dev -- ...  # 通过 tsx 从源码运行
+```
+
+每个智能体是 [`src/targets/`](./src/targets) 下的一个适配器，实现
+`apply` / `prune` / `current` / `candidates`；只读的导入来源放在
+[`src/sources/`](./src/sources)。新增一个应用 = 一个文件 +
+[`src/targets/index.ts`](./src/targets/index.ts) 里的一行。
+
+## 许可证
+
+MIT

package/dist/apps.js

@@ -0,0 +1,178 @@
+import { execFileSync, execSync } from "node:child_process";
+import fs from "node:fs";
+import path from "node:path";
+import { home, readJsonIfExists } from "./fsutil.js";
+export const appPackages = [
+    {
+        id: "claude",
+        name: "Claude Code",
+        binary: "claude",
+        latest: { kind: "npm", name: "@anthropic-ai/claude-code" },
+        installCmd: "curl -fsSL https://claude.ai/install.sh | bash",
+        upgradeCmd: "claude update",
+    },
+    {
+        id: "codex",
+        name: "Codex CLI",
+        binary: "codex",
+        latest: { kind: "npm", name: "@openai/codex" },
+        installCmd: "npm install -g @openai/codex@latest",
+    },
+    {
+        id: "omp",
+        name: "Oh My Pi",
+        binary: "omp",
+        latest: { kind: "brew", name: "omp" },
+        installCmd: "brew install omp",
+        upgradeCmd: "brew upgrade omp",
+    },
+    {
+        id: "pi",
+        name: "pi",
+        binary: "pi",
+        latest: { kind: "npm", name: "@earendil-works/pi-coding-agent" },
+        installCmd: "npm install -g @earendil-works/pi-coding-agent@latest",
+    },
+    {
+        id: "prime",
+        name: "prime-agent",
+        binary: "prime-agent",
+        latest: { kind: "github", name: "PrimeIntellect-ai/prime-agent" },
+        installCmd: "curl -fsSL https://raw.githubusercontent.com/PrimeIntellect-ai/prime-agent/main/install.sh | bash",
+    },
+    {
+        id: "opencode",
+        name: "opencode",
+        binary: "opencode",
+        latest: { kind: "npm", name: "opencode-ai" },
+        installCmd: "curl -fsSL https://opencode.ai/install | bash",
+        upgradeCmd: "opencode upgrade",
+    },
+    {
+        id: "hermes",
+        name: "Hermes",
+        binary: "hermes",
+        latest: { kind: "pypi", name: "hermes-agent" },
+        installCmd: "uv tool install hermes-agent || pipx install hermes-agent",
+        upgradeCmd: "uv tool upgrade hermes-agent || pipx upgrade hermes-agent",
+    },
+    {
+        id: "workbuddy",
+        name: "WorkBuddy",
+        // Electron desktop app: version from its data dir; managed by its own updater.
+        localVersion: () => {
+            const j = readJsonIfExists(path.join(home, ".workbuddy", "last-launch.json"));
+            const v = (j?.version ?? j?.appVersion);
+            if (v)
+                return v;
+            const r = readJsonIfExists(path.join(home, ".workbuddy", "app", "renderer-version.json"));
+            return r?.version ?? undefined;
+        },
+    },
+    {
+        id: "dsh",
+        name: "DeepSeek Harness",
+        binary: "dsh",
+        latest: { kind: "npm", name: "@deepseek-ai/dsh" },
+        installCmd: "npm install -g @deepseek-ai/dsh@latest",
+    },
+];
+const SEMVERISH = /\d+\.\d+(\.\d+)?([-.+][\w.-]+)?/;
+export function installedVersion(app) {
+    if (app.binary) {
+        try {
+            const out = execFileSync(app.binary, app.versionArgs ?? ["--version"], {
+                encoding: "utf8",
+                timeout: 15000,
+                stdio: ["ignore", "pipe", "pipe"],
+            });
+            return SEMVERISH.exec(out)?.[0];
+        }
+        catch {
+            // binary present but probe failed (e.g. runtime version gate) -> installed, version unknown
+            if (binaryOnPath(app.binary))
+                return "?";
+            return app.localVersion?.();
+        }
+    }
+    return app.localVersion?.();
+}
+async function fetchJson(url, headers) {
+    const res = await fetch(url, { headers, signal: AbortSignal.timeout(15000) });
+    if (!res.ok)
+        throw new Error(`${url} -> ${res.status}`);
+    return res.json();
+}
+export async function latestVersion(app) {
+    const src = app.latest;
+    if (!src)
+        return undefined;
+    try {
+        switch (src.kind) {
+            case "npm": {
+                const data = (await fetchJson(`https://registry.npmjs.org/${src.name}/latest`));
+                return data.version;
+            }
+            case "pypi": {
+                const data = (await fetchJson(`https://pypi.org/pypi/${src.name}/json`));
+                return data.info?.version;
+            }
+            case "github": {
+                const data = (await fetchJson(`https://api.github.com/repos/${src.name}/releases/latest`, {
+                    "user-agent": "agentsw",
+                }));
+                return data.tag_name ? (SEMVERISH.exec(data.tag_name)?.[0] ?? data.tag_name) : undefined;
+            }
+            case "brew": {
+                // Local tap metadata first (works for taps; refreshed by `brew update`), then core API.
+                try {
+                    const out = execSync(`brew info --json=v2 ${src.name}`, { encoding: "utf8", timeout: 30000 });
+                    const data = JSON.parse(out);
+                    const stable = data.formulae?.[0]?.versions?.stable;
+                    if (stable)
+                        return stable;
+                }
+                catch {
+                    /* brew missing or formula unknown locally */
+                }
+                const data = (await fetchJson(`https://formulae.brew.sh/api/formula/${src.name}.json`));
+                return data.versions?.stable;
+            }
+        }
+    }
+    catch {
+        return undefined;
+    }
+}
+/** numeric-aware semver-ish comparison; true when b is newer than a */
+export function isNewer(installed, latest) {
+    const pa = installed.split(/[.+-]/).map(Number);
+    const pb = latest.split(/[.+-]/).map(Number);
+    for (let i = 0; i < Math.max(pa.length, pb.length); i++) {
+        const a = pa[i] ?? 0;
+        const b = pb[i] ?? 0;
+        if (Number.isNaN(a) || Number.isNaN(b))
+            break;
+        if (b > a)
+            return true;
+        if (b < a)
+            return false;
+    }
+    return false;
+}
+/** Run an install/upgrade command with live output. Throws on nonzero exit. */
+export function runShell(command) {
+    execSync(command, { stdio: "inherit", env: process.env });
+}
+export function binaryOnPath(binary) {
+    const dirs = (process.env.PATH ?? "").split(path.delimiter);
+    return dirs.some((d) => {
+        try {
+            fs.accessSync(path.join(d, binary), fs.constants.X_OK);
+            return true;
+        }
+        catch {
+            return false;
+        }
+    });
+}