> 译自[英文 README](../README.md)。以英文版为准。 # apfel ### 你的 Mac 上早已内置的免费 AI。 [![Swift 6.3+](https://img.shields.io/badge/Swift-6.3%2B-F05138?logo=swift&logoColor=white)](https://swift.org) [![macOS 26 Tahoe+](https://img.shields.io/badge/macOS-26%20Tahoe%2B-000000?logo=apple&logoColor=white)](https://developer.apple.com/macos/) [![No Xcode Required](https://img.shields.io/badge/Xcode-not%20required-orange)](https://developer.apple.com/xcode/resources/) [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](../LICENSE) [![100% On-Device](https://img.shields.io/badge/inference-100%25%20on--device-green)](https://developer.apple.com/documentation/foundationmodels) [![Website](https://img.shields.io/badge/web-apfel.franzai.com-16A34A)](https://apfel.franzai.com) [![#agentswelcome](https://img.shields.io/badge/%23agentswelcome-PRs%20welcome-0066cc?style=for-the-badge&labelColor=0d1117&logo=probot&logoColor=white)](../README.md#contributing) Apple Silicon Mac 通过 [Apple FoundationModels](https://developer.apple.com/documentation/foundationmodels) 内置了一个大语言模型。`apfel` 把它封装成 UNIX 工具和本地的 OpenAI 兼容服务器。100% 在设备端运行。无需 API 密钥,无需云端。 | 模式 | 命令 | 你能得到什么 | |------|---------|--------------| | UNIX 工具 | `apfel "prompt"` / `echo "text" \| apfel` | 适合管道的回答、文件附件、JSON 输出、退出码 | | OpenAI 兼容服务器 | `apfel --serve` | 可直接替换的本地 `http://localhost:11434/v1` 后端,供 OpenAI SDK 使用 | `apfel --chat` - 交互式 REPL。 [![apfel 终端演示视频: 真实命令, 设备端真实回答 (82 秒)](../screenshots/demo-video.jpg)](https://apfel.franzai.com/zh-hans/a/apfel-ai-tutorial/#video) 所有场景下均支持工具调用。 ![apfel CLI](../screenshots/cli.png) ## 系统要求与安装 macOS 26 Tahoe+,Apple Silicon(M1+),[已启用 Apple Intelligence](https://support.apple.com/en-us/121115)。在 macOS 27 上,搭载 M3 或更新芯片且内存 12 GB 以上的 Mac 会运行 Apple 更大的设备端模型,上下文窗口为 8192 个 token;M1/M2 Mac 以及所有运行 macOS 26 的 Mac 为 4096。`apfel --model-info` 会显示你的实际值。 ```bash brew install apfel ``` 更新: ```bash brew upgrade apfel ``` 从源码构建(使用带 macOS 27 SDK / Swift 6.3 的命令行工具,无需 Xcode;生成的二进制文件可在 macOS 26 和 27 上运行): ```bash git clone https://github.com/Arthur-Ficial/apfel.git && cd apfel && make install ``` 当日更新的 tap、Mint、mise、故障排查:[docs/install.md](../docs/install.md)。 **Languages:** [English](../README.md) | [العربية](README.ar.md) | [Беларуская](README.be.md) | [Български](README.bg.md) | [বাংলা](README.bn.md) | [Bosanski](README.bs.md) | [Català](README.ca.md) | [Čeština](README.cs.md) | [Dansk](README.da.md) | [Deutsch](README.de.md) | [Ελληνικά](README.el.md) | [Español](README.es.md) | [Eesti](README.et.md) | [فارسی](README.fa.md) | [Suomi](README.fi.md) | [Français](README.fr.md) | [ગુજરાતી](README.gu.md) | [हिन्दी](README.hi.md) | [Hrvatski](README.hr.md) | [Magyar](README.hu.md) | [Italiano](README.it.md) | [日本語](README.ja.md) | [ಕನ್ನಡ](README.kn.md) | [한국어](README.ko.md) | [Lietuvių](README.lt.md) | [Latviešu](README.lv.md) | [Македонски](README.mk.md) | [മലയാളം](README.ml.md) | [मराठी](README.mr.md) | [Norsk bokmål](README.nb.md) | [Nederlands](README.nl.md) | [Polski](README.pl.md) | [Português](README.pt.md) | [Română](README.ro.md) | [Русский](README.ru.md) | [Slovenčina](README.sk.md) | [Slovenščina](README.sl.md) | [Shqip](README.sq.md) | [Srpski](README.sr.md) | [Svenska](README.sv.md) | [தமிழ்](README.ta.md) | [తెలుగు](README.te.md) | [Türkçe](README.tr.md) | [Українська](README.uk.md) | [اردو](README.ur.md) | **简体中文** | [繁體中文](README.zh-hant.md) ## 快速开始 ### UNIX 工具 含有 `!` 的 prompt 请用单引号括起来(zsh/bash 的历史记录展开):`apfel 'Hello, Mac!'`。 ```bash # Single prompt apfel "What is the capital of Austria?" # Permissive mode - reduces guardrail false positives for creative/long prompts apfel --permissive "Write a dramatic opening for a thriller novel" # Stream output apfel --stream "Write a haiku about code" # Pipe input echo "Summarize: $(cat README.md)" | apfel # Attach file content to prompt apfel -f README.md "Summarize this project" # Attach multiple files apfel -f old.swift -f new.swift "What changed between these two files?" # Attach a PDF or image - on-device text extraction, OCR, and "what the image is about" apfel -f report.pdf "Summarize the key findings" apfel -f receipt.jpg "What is the total?" # Pipe a file straight in (PDF, image, or text) cat report.pdf | apfel "Summarize this" # Combine files with piped input git diff HEAD~1 | apfel -f CONVENTIONS.md "Review this diff against our conventions" # Only the code - no prose, no markdown fences (pipe-safe, exit 7 if empty) apfel --code "a python function that deduplicates a list" > dedupe.py apfel --code "shell one-liner to find the 10 largest files here" | pbcopy # JSON output for scripting apfel -o json "Translate to German: hello" | jq .content # Guaranteed schema-valid JSON output (guided generation) apfel --schema person.schema.json "Extract the person: Alice is 30." | jq .name # One-shot multi-turn: conversation JSON in, next assistant turn out jq '. += [{"role":"user","content":"and in German?"}]' conv.json | apfel --messages - # Batch: JSONL records in, one JSON result line per record out (fresh session each) jq -c '{custom_id:(.id|tostring),prompt:("Classify as billing, bug, or question: " + .text)}' tickets.jsonl \ | apfel --batch | jq -c 'select(.status=="ok") | {id:.custom_id,category:.content}' # Preflight token budget before a large prompt apfel --count-tokens -f README.md "Summarize this" # System prompt apfel -s "You are a pirate" "What is recursion?" # System prompt from file apfel --system-file persona.txt "Explain TCP/IP" # Quiet mode for shell scripts result=$(apfel -q "Capital of France? One word.") ``` ### OpenAI 兼容服务器 ```bash apfel --serve # foreground brew services start apfel # background (like Ollama) brew services stop apfel APFEL_TOKEN=$(uuidgen) APFEL_MCP=/path/to/tools.py brew services start apfel ``` ```bash curl http://localhost:11434/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{"model":"apple-foundationmodel","messages":[{"role":"user","content":"Hello"}]}' ``` ```python from openai import OpenAI client = OpenAI(base_url="http://localhost:11434/v1", api_key="unused") resp = client.chat.completions.create( model="apple-foundationmodel", messages=[{"role": "user", "content": "What is 1+1?"}], ) print(resp.choices[0].message.content) ``` 后台服务详情:[docs/background-service.md](../docs/background-service.md)。 ### 快速测试聊天 `apfel --chat` 是一个小型 REPL,用于测试 prompt 或 MCP 服务器。如果需要图形界面的聊天应用,请参阅 [apfel-chat](https://github.com/Arthur-Ficial/apfel-chat)。 ```bash apfel --chat apfel --chat -s "You are a helpful coding assistant" apfel --chat --mcp ./mcp/calculator/server.py # chat with MCP tools apfel --chat --debug # debug output to stderr ``` 按 Ctrl-C 退出。上下文会被自动裁剪([docs/context-strategies.md](../docs/context-strategies.md))。 在基于它编写脚本之前,值得了解的限制(各 Mac 的上下文窗口、安全护栏、不支持 embeddings):[限制](../README.md#limitations)。 ## 演示 封装了 `apfel` 的真实 shell 脚本:`cmd`(英语转 shell 命令)、`oneliner`、`mac-narrator`、`wtd`、`explain`、`naming`、`port`、`gitsum`。它们已内置在二进制文件中,无需克隆仓库。把它们写到你喜欢的位置即可,无论你是从 homebrew-core、tap 还是源码安装的 apfel: ```bash apfel demos ./apfel-demos ``` 这会把每个演示(可执行文件)以及一个 `README.md` 写入 `./apfel-demos`。然后从那里运行它们: ```bash ./apfel-demos/cmd "find all .log files modified today" # $ find . -name "*.log" -type f -mtime -1 ./apfel-demos/cmd -x "show disk usage sorted by size" # -x = execute after confirm ./apfel-demos/cmd -c "list open ports" # -c = copy to clipboard ``` 执行 `brew upgrade apfel` 之后,请重新运行 `apfel demos` 以刷新。源码:[`demo/`](../demo/)。详细演练以及可复制粘贴到 `.zshrc` 的 `cmd` shell 函数:[docs/demos.md](../docs/demos.md)。 ## MCP 工具支持 用 `--mcp` 挂载 [Model Context Protocol](https://modelcontextprotocol.io/) 服务器。apfel 负责发现、调用并返回结果。 ```bash apfel --mcp ./mcp/calculator/server.py "What is 15 times 27?" ``` ``` mcp: ./mcp/calculator/server.py - add, subtract, multiply, divide, sqrt, power, round_number ← stderr tool: multiply({"a": 15, "b": 27}) = 405 ← stderr 15 times 27 is 405. ← stdout ``` 使用 `-q` 可隐藏工具信息。 ```bash apfel --mcp ./server_a.py --mcp ./server_b.py "Use both tools" apfel --serve --mcp ./mcp/calculator/server.py apfel --chat --mcp ./mcp/calculator/server.py ``` 自带一个计算器,位于 [`mcp/calculator/`](../mcp/calculator/)([docs/mcp-calculator.md](../docs/mcp-calculator.md))。 **远程 MCP 服务器**(Streamable HTTP,MCP 规范 2025-03-26): ```bash apfel --mcp https://mcp.example.com/v1 "what tools do you have?" # bearer token - prefer env var (flag is visible in ps aux) APFEL_MCP_TOKEN=mytoken apfel --mcp https://mcp.example.com/v1 "..." # mixed local + remote apfel --mcp /path/to/local.py --mcp https://remote.example.com/v1 "..." ``` > **安全提示:** 请优先使用 `APFEL_MCP_TOKEN` 而不是 `--mcp-token`(后者会出现在 ps aux 中)。apfel 拒绝通过明文 `http://` 发送 bearer token。 ## apfel-run:可选的配置层 apfel 本身没有配置文件 - 和任何 UNIX 工具一样,只用命令行参数和环境变量。如果你想要 TOML 配置(多个 MCP、配置档案、在 git 中管理团队配置),可以使用 [**apfel-run**](https://github.com/Arthur-Ficial/apfel-run),这是一个 MIT 许可的封装工具,通过 `execve` 以即插即用的方式提供配置。 ```bash brew install Arthur-Ficial/tap/apfel-run apfel-run config init # starter ~/.config/apfel/config.toml alias apfel=apfel-run # optional, every apfel flag still works ``` ## OpenAI API 兼容性 **Base URL:** `http://localhost:11434/v1` | 功能 | 状态 | 说明 | |---------|--------|-------| | `POST /v1/chat/completions` | 支持 | 流式 + 非流式 | | `POST /v1/responses` | 支持 | OpenAI Responses API:字符串/消息输入、`instructions`、流式(标准事件序列)、`text.format`(包括 `json_schema`)、函数工具(非流式)。有状态的功能(`previous_response_id`、`store: true`、`background`、`reasoning`、托管工具)会如实返回 501 | | `POST /v1/responses/input_tokens` | 支持 | Token 预检(`client.responses.input_tokens.count(...)`):由 Apple 的分词器给出精确计数,不进行生成;低于 macOS 26.4 时返回 501,而不是给出估算值。参见 [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md#token-preflight-v1responsesinput_tokens) | | `GET /v1/models` | 支持 | 返回 `apple-foundationmodel`(通用,默认)和 `apple-foundationmodel-content-tagging`(标签,非散文)。参见 [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md#content-tagging-model) | | `GET /health` | 支持 | 模型可用性、上下文窗口、语言、上报的模型能力(`capabilities`、`capabilities_reported`)、当前服务的模型变体(`variant`,例如 AFM 3 Core) | | `GET /v1/logs`, `/v1/logs/stats` | 仅调试 | 需要 `--debug` | | 工具调用 | 支持 | 调用通过引导式生成写出(`--tool-calling guided`,默认),或以服务器解析的 JSON 形式写出(`--tool-calling text`)。参见 [docs/tool-calling-guide.md](../docs/tool-calling-guide.md) | | `response_format: json_object` | 支持 | 通过注入 system prompt 实现;输出中的 markdown 代码围栏会被去除 | | `response_format: json_schema` | 支持 | 通过 FoundationModels `DynamicGenerationSchema` 保证输出符合 schema;可与 `stream: true` 配合使用 | | `temperature`, `top_p`, `max_tokens`, `seed` | 支持 | 映射到 `GenerationOptions`。`top_p` 为核采样;`temperature` 低于 0.001(含 0)时映射为贪心解码(确定性)。省略 `max_tokens` 时使用剩余的上下文窗口(与 OpenAI 的语义完全一致)- 参见[默认响应上限](../README.md#default-response-cap-max_tokens) | | `stream: true` | 支持 | SSE;仅当设置 `stream_options: {"include_usage": true}` 时才发送最后的 usage 数据块(遵循 OpenAI 规范) | | `usage` | 支持 | 在 macOS 26.4+ 上为真实的分词器计数(更低版本为 chars/4 估算)。macOS 27:使用运行时自身的统计,包含 Apple 聊天模板的额外开销(单行 prompt 的 `prompt_tokens` 约多 46);macOS 26:apfel 统计消息文本。详情见 [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md) | | `finish_reason` | 支持 | `stop`、`tool_calls`、`length` | | 上下文策略 | 支持 | 扩展字段 `x_context_strategy`、`x_context_max_turns`、`x_context_output_reserve` | | CORS | 支持 | 使用 `--cors` 启用 | | `POST /v1/completions` | 501 | 不支持旧版文本补全 | | `POST /v1/embeddings` | 501 | 设备端不提供 embeddings | | `stop` | 支持 | 字符串或最多 4 个字符串;仅限纯文本(在有工具参与或使用 `json_object`/`json_schema` 时返回 400)。在输出端应用,流式与非流式行为一致。参见 [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md#stop-sequences) | | `logprobs=true`, `n>1`, `presence_penalty`, `frequency_penalty` | 400 | 明确拒绝。`n=1` 和 `logprobs=false` 会被接受,但不起任何作用 | | 多模态(图像) | macOS 27 上支持 | 以 base64 data URL 形式传入的 `image_url` 部分会原生交给设备端模型(base64 上限 20 MB,最长边会缩小到 4096 px)。远程 `http(s)` URL 和文件路径一律返回 400 - apfel 绝不会抓取图像,也绝不会为 HTTP 客户端读取文件。在 macOS 26 上:返回 400 及明确的错误信息。参见 [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md) | | `Authorization` 请求头 | 支持 | 设置 `--token` 后为必需。参见 [docs/server-security.md](../docs/server-security.md) | 完整 API 规范:[openai/openai-openapi](https://github.com/openai/openai-openapi)。 ## 默认响应上限(`max_tokens`) 省略 `max_tokens` 时,**CLI 与 OpenAI 兼容服务器的行为完全一致**:该值以 `nil` 传入,模型会使用上下文窗口中剩余的全部空间。这与 OpenAI 的语义完全一致,没有任意的回退常量。 设备端上下文窗口同时容纳输入*和*输出:**macOS 26 以及 M1/M2 Mac 上为 4096 个 token,搭载 M3 或更新芯片且内存 12 GB 以上、运行 macOS 27 的 Mac 上为 8192 个 token**。apfel 在运行时通过 `SystemLanguageModel.contextSize` 读取真实大小 - 用 `apfel --model-info` 查看你的实际值。如果生成触及上限,响应会以 `finish_reason: "length"` 正常结束,并返回已生成的部分内容(服务器:HTTP 200;CLI:退出码 0 并在 stderr 给出警告,使用 `--require-complete` 时退出码为 8 - 参见 [docs/cli-reference.md](../docs/cli-reference.md))。如果你想要更严格的延迟预算,或为客户端设定已知的上限,请显式传入 `max_tokens`。 ### 示例 ```bash # Omitted: uses remaining window, finish_reason: "stop" or "length" curl -sS http://localhost:11434/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{"model":"apple-foundationmodel", "messages":[{"role":"user","content":"Reply SKIP, MOVE, or RENAME."}]}' # Explicit cap (recommended for tight latency budgets) curl -sS http://localhost:11434/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{"model":"apple-foundationmodel","max_tokens":128, "messages":[{"role":"user","content":"Summarise: ..."}]}' ``` ### 如何选值 | 使用场景 | `max_tokens` | |----------------------------------------|---------------| | 单词 / 分类回复 | 16 - 32 | | 一行指令 | 64 - 128 | | 短段落 | 256 - 512 | | 长段落 / 结构化 JSON | 1024 - 2048 | | 尽可能占满上下文窗口 | 省略 | 请让 `input_tokens + max_tokens` 保持在上下文窗口以内并留有余量(macOS 26 以及 M1/M2 Mac 上为 4096 个 token,搭载 M3 或更新芯片且内存 12 GB 以上、运行 macOS 27 的 Mac 上为 8192(Apple 的 AFM 3 Core Advanced 模型))。如果 prompt 本身就超出窗口,生成将无法开始,请求会以 `[context overflow]` 失败(HTTP 400 / CLI 退出码 4)。校验器会拒绝非正数值(`max_tokens <= 0`)。 ### CLI 一致性 CLI 与服务器遵循同一条规则:省略 = 使用剩余窗口。没有会漂移的常量。可用 `--max-tokens N` 或 `APFEL_MAX_TOKENS=N` 覆盖。 ```bash apfel "Reply SKIP." # uses remaining window apfel --max-tokens 64 "Reply SKIP." # explicit cap APFEL_MAX_TOKENS=2048 apfel "..." # via env var ``` ### 服务器的宽松安全护栏 `apfel --serve --permissive` 会让服务器对该进程处理的**每个请求**使用 Apple 的 `.permissiveContentTransformations` 安全护栏。标志相同,语义也与 CLI 的 `--permissive` 相同([docs/PERMISSIVE.md](../docs/PERMISSIVE.md))。不支持按请求覆盖 - 由服务器运维者对整个进程做出决定。 ```bash apfel --serve --permissive # every request uses permissive guardrails ``` ## 限制 | 约束 | 详情 | |------------|--------| | 上下文窗口 | **macOS 26 以及 M1/M2 Mac 上为 4096 个 token,搭载 M3 或更新芯片且内存 12 GB 以上、运行 macOS 27 的 Mac 上为 8192 个 token**(输入 + 输出合计)。并非硬编码常量 - apfel 在运行时读取 `SystemLanguageModel.contextSize`;`apfel --model-info` 会打印当前的实际值 | | 平台 | 仅限 macOS 26+ 和 Apple Silicon | | 模型 | 一个设备端模型(约 3B 参数),以 `apple-foundationmodel`(通用)和 `apple-foundationmodel-content-tagging`(Apple 的内容标记适配器;不支持工具)两种身份提供 | | 安全护栏 | Apple 的安全系统可能会拦截无害的 prompt。`--permissive` 可减少误报([docs/PERMISSIVE.md](../docs/PERMISSIVE.md)) | | 速度 | 在设备端运行,不是云端规模 - 每次响应需要几秒钟 | | 无 embeddings | 设备端不提供 | | 视觉 | 图像输入仅在 macOS 27 上才会到达模型(服务器:data URL;CLI:`-f photo.jpg` 以及通过管道传入的图像,在 OCR 文本之外附加)。在 macOS 26 上,图像仅以文本形式通过 OCR/分类处理。3B 模型对场景的描述较为粗略 - 要读取印刷文字,OCR 仍是可靠的渠道 | | 训练数据 / 知识截止日期 | Apple 没有公布设备端模型的确切截止日期。被追问时,模型**每次采样都会虚构一个不同的日期**(例如"2023 年 10 月"、"2023 年 4 月")。请把模型对自身训练情况的所有自述都视为不可靠。 | | 不知道当前日期,也没有实时感知 | 模型不知道今天的日期,也无法访问网络或时钟。被问到时,它要么拒绝回答,要么编造一个日期。当计算依赖当前日期时,请通过 system prompt 注入(见下方的变通办法)。 | **依赖日期的 prompt 的变通办法** - 以 system 消息的形式注入当前日期: ```bash apfel -s "Today is $(date '+%B %d, %Y')." "Write a one-line release note dated today." ``` ```bash apfel --chat -s "Today is $(date '+%B %d, %Y'). You are a helpful assistant." ``` 注意:即使注入了日期,3B 模型仍可能产生幻觉(尤其是被直接问及自身训练截止日期时)。注入日期对*使用*该日期的生成类 prompt 有帮助,但无法改变模型"自述"的反射习惯。 背景:[#158](https://github.com/Arthur-Ficial/apfel/issues/158)。 ## 参考文档 从 [Python](../docs/guides/python.md)、[Node.js](../docs/guides/nodejs.md)、[Ruby](../docs/guides/ruby.md)、[PHP](../docs/guides/php.md)、[Bash/curl](../docs/guides/bash-curl.md)、[Zsh](../docs/guides/zsh.md)、[AppleScript](../docs/guides/applescript.md)、[Swift](../docs/guides/swift-scripting.md)、[Perl](../docs/guides/perl.md)、[AWK](../docs/guides/awk.md) 使用 apfel 的指南 - 参见 [docs/guides/index.md](../docs/guides/index.md)。均经过实测;可运行的验证见 [apfel-guides-lab](https://github.com/Arthur-Ficial/apfel-guides-lab)。 - [docs/install.md](../docs/install.md) - 安装、故障排查和 Apple Intelligence 设置 - [docs/cli-reference.md](../docs/cli-reference.md) - 所有标志、退出码和环境变量 - [docs/file-extraction.md](../docs/file-extraction.md) - `-f` 以及通过管道传入的 PDF / 图像 / 文本提取(OCR + 图像理解) - [docs/background-service.md](../docs/background-service.md) - `brew services` 和 launchd 的用法 - [docs/openai-api-compatibility.md](../docs/openai-api-compatibility.md) - `/v1/*` 支持矩阵详解 - [docs/server-security.md](../docs/server-security.md) - 来源检查、CORS、token 和 `--footgun` - [docs/context-strategies.md](../docs/context-strategies.md) - 聊天裁剪策略 - [docs/mcp-calculator.md](../docs/mcp-calculator.md) - 本地与远程 MCP 的用法 - [docs/tool-calling-guide.md](../docs/tool-calling-guide.md) - 工具调用行为详解 - [docs/integrations.md](../docs/integrations.md) - 第三方工具集成(opencode 等) - [docs/local-setup-with-vs-code.md](../docs/local-setup-with-vs-code.md) - 在 VS Code 中使用 apfel + 另一个编辑/应用模型进行本地审查 - [docs/demos.md](../docs/demos.md) - shell 演示的更长篇演练 - [docs/EXAMPLES.md](../docs/EXAMPLES.md) - 50 多个真实 prompt 及未经编辑的输出 - [docs/swift-library.md](../docs/swift-library.md) - 面向下游开发者的 `ApfelCore` Swift Package - [docs/coreai-impact.md](../docs/coreai-impact.md) - 为什么 apfel 基于 FoundationModels 而不是 Apple 新的 Core AI(Core ML 的继任者) - [docs/apfel-vs-fm.md](../docs/apfel-vs-fm.md) - apfel 与 Apple 自家 `fm` CLI(macOS 27)的逐项功能对比及实测数据 ## 架构 ```text CLI (single/stream/chat) ──┐ ├─→ FoundationModels.SystemLanguageModel HTTP Server (/v1/*) ───────┘ (100% on-device, zero network) ContextManager → Transcript API SchemaConverter → native ToolDefinitions TokenCounter → real token counts (SDK 26.4) ``` Swift 6.3 严格并发。三个 target:`ApfelCore`(纯逻辑,可单元测试,也作为 Swift Package 产品提供 - 链接见上方"参考文档")、`apfel`(CLI + 服务器)以及 `apfel-tests`(纯 Swift 运行器,不依赖 XCTest)。 ## 构建与测试 ```bash make test # release build + all unit/integration tests make preflight # full release qualification make install # build release + install to /usr/local/bin make build # build release only make version # print current version make release # patch release make release TYPE=minor # minor release make release TYPE=major # major release swift build # quick debug build (no version bump) swift run apfel-tests # unit tests python3 -m pytest Tests/integration/ -v # integration tests apfel --benchmark -o json # performance report ``` `.version` 是唯一的事实来源。只有 `make release` 会提升版本号。本地构建不会改变版本。 ## apfel 生态树 基于 apfel 构建的项目。每个项目都有自己的仓库和 Homebrew formula。 | 项目 | 功能 | 安装 | |---------|--------------|---------| | [**apfel**](https://apfel.franzai.com) | 根基。设备端 FoundationModels CLI + OpenAI 兼容服务器。 | `brew install apfel` | | [**apfel-chat**](https://apfel-chat.franzai.com) | macOS 聊天客户端:流式 markdown、语音输入输出、Apple Vision 图像分析。 | `brew install Arthur-Ficial/tap/apfel-chat` | | [**apfel-clip**](https://apfel-clip.franzai.com) | 菜单栏中针对剪贴板的 AI 操作:总结、翻译、改写。 | `brew install Arthur-Ficial/tap/apfel-clip` | | [**apfel-quick**](https://apfel-quick.franzai.com) | 即时 AI 浮层:按下快捷键,提问,得到答案,关闭。 | `brew install Arthur-Ficial/tap/apfel-quick` | | [**apfelpad**](https://apfelpad.franzai.com) | 公式记事本 - 设备端 AI 作为内联单元格函数。 | `brew install Arthur-Ficial/tap/apfelpad` | | [**apfel-mcp**](https://apfel-mcp.franzai.com) | 针对小型设备端窗口优化 token 预算的 MCP:`url-fetch`、`ddg-search`、`search-and-fetch`。 | `brew install Arthur-Ficial/tap/apfel-mcp` | | [**apfel-gui**](https://github.com/Arthur-Ficial/apfel-gui) | SwiftUI 调试检查器:请求时间线、MCP 协议查看器、TTS/STT。 | `brew install Arthur-Ficial/tap/apfel-gui` | | [**apfel-run**](https://github.com/Arthur-Ficial/apfel-run) | UNIX 封装工具,在 `apfel` 之上增加持久化的 MCP 注册表和 TOML 配置。 | `brew install Arthur-Ficial/tap/apfel-run` | | [**apfel-tag**](https://github.com/Arthur-Ficial/apfel-tag) | 设备端内容标注 CLI:通过管道输入文本,输出标签/主题/情绪。 | `brew install Arthur-Ficial/tap/apfel-tag` | | [**apfel-server-kit**](https://github.com/Arthur-Ficial/apfel-server-kit) | 面向生态工具的 Swift package:发现、启动本地 `apfel --serve` 并从中获取流式输出。 | Swift Package | ## 社区项目 基于 apfel 做出了东西?提交一个 issue,就可以把它加到这里。 | 项目 | 功能 | 链接 | |---------|-------------|-------| | **apfelclaw**,作者 [@julianYaman](https://github.com/julianYaman) | 本地 AI 智能体,通过只读工具读取文件、日历、邮件和 Mac 状态 | [github](https://github.com/julianyaman/apfelclaw) - [site](https://apfelclaw.yamanlabs.com/) | | **fruit-chat**,作者 [@bhaskarvilles](https://github.com/bhaskarvilles) | 基于浏览器的聊天界面,通过 OpenAI 兼容 API 与 `apfel --serve` 通信 | [github](https://github.com/bhaskarvilles/fruit-chat) | | **local-claude**,作者 [@lucaspwo](https://github.com/lucaspwo) | Claude Code 封装工具,通过一个小型 Anthropic-OpenAI 代理把 apfel 换成本地后端 | [github](https://github.com/lucaspwo/local-claude) | | **apfeller**,作者 [@hasit](https://github.com/hasit) | 围绕 apfel 构建的本地 shell 应用的应用管理器 | [github](https://github.com/hasit/apfeller) - [site](https://hasit.github.io/apfeller/) - [catalog](https://hasit.github.io/apfeller/catalog/) | | **apfel-for-raycast**,作者 [@eggsy](https://github.com/eggsy) | Raycast 命令栏扩展:提问、翻译、解释文件和目录、对话历史、自定义 system prompt。通过 apfel CLI 在设备端运行。 | [store](https://www.raycast.com/eggsy/apfel) - [github](https://github.com/raycast/extensions/tree/main/extensions/apfel) | ## 参与贡献 欢迎在任何 `Arthur-Ficial/apfel*` 仓库提交 issue 和 PR。 **#agentswelcome** - 欢迎 AI 智能体提交 PR。请阅读仓库中的 `CLAUDE.md`,运行测试,并在 `Co-Authored-By` trailer 中注明所用工具。标准与人类相同:代码整洁、测试通过、如实说明限制。对智能体最友好的入口:[apfel-mcp](https://github.com/Arthur-Ficial/apfel-mcp)([贡献规则](https://apfel-mcp.franzai.com/#contribute))。 ## 许可证 [MIT](../LICENSE)