omni-context-cli 0.0.103 → 0.0.105

package/README.md

@@ -1,6 +1,6 @@
 # OmniContext CLI
 
-A context-first, zero-telemetry coding assistant that extends into VS Code, Office, the browser, and mobile. Supports Anthropic, OpenAI, Gemini, and Responses API, plus chat image generation through Google Imagen, OpenAI Images, and xAI Images, and Image Workshop image/video providers for Google Imagen, OpenAI Images, Google Videos, and xAI, with custom workflows, atlas navigation, and cross-session memory.
+A context-first, zero-telemetry coding assistant that extends into VS Code, Office, the browser, and mobile. Supports Anthropic, OpenAI, Gemini, and the Responses API, plus chat image generation through Google Imagen, OpenAI Images, and xAI Images, and Image Workshop image/video providers for Google Imagen, OpenAI Images, Google Videos, and xAI, with custom workflows, atlas navigation, and cross-session memory.
 
 ## Context-First Architecture
 
@@ -9,7 +9,7 @@ Every token matters. OmniContext CLI is engineered from the ground up to squeeze
 - **Lean system prompts** - the built-in system prompts are short, focused, and free of boilerplate. Tool descriptions are minimal. Your context budget goes toward actual work, not framework overhead.
 - **Agentic tools** - exploratory tasks (searching code, surveying project structure, previewing files) run as separate agentic tools, optionally on a dedicated auxiliary model. Intermediate tool calls, file contents, and reasoning never enter your main context. You get a concise answer back; the scratch work stays off the books.
 - **Atlas navigation** - code navigation knowledge is stored as lightweight topic-to-file maps, so assistants can jump to the right files before searching from scratch.
-- **Cross-session memory** - every time a conversation is compacted, the model distills key points and evaluates existing memories. Good insights accumulate weight, stale knowledge decays naturally. When the next session starts, accumulated memories are injected automatically so the assistant knows your preferences and project context from the first message.
+- **Cross-session memory** - every time a conversation is compacted, the model distills key points and evaluates existing memories. Good insights accumulate weight, stale knowledge decays naturally. When a new chat session starts, accumulated memories are injected automatically so the assistant knows your preferences and project context from the first message.
 - **Prompt caching** - you choose the cache TTL: 5 minutes for quick iteration, 1 hour for long work sessions. Cache control for Anthropic is applied automatically, so repeated context blocks are served from cache, cutting token usage and cost significantly.
 - **Context editing** - for long, tool-heavy sessions, an advanced setting can compress older tool inputs, tool outputs, and reasoning traces before the next request to reduce token usage while keeping saved history intact.
 
@@ -61,7 +61,7 @@ Each LLM API protocol has its own dedicated request builder and stream handler.
 | **Google Videos** | Google video generation for Image Workshop |
 | **xAI Videos** | xAI video generation API for Image Workshop |
 
-Provider-specific interceptors handle protocol and model-source quirks such as Anthropic next-gen/legacy models, CPA proxy session affinity, DeepSeek, Responses API adjustments and service tier handling, Gemini legacy models, Kimi, MiniMax, OpenCode Go, generic OpenAI-compatible endpoints, OpenRouter metadata and caching, xAI, Zhipu GLM, and Zenmux model routing without compromising the protocol.
+Provider-specific interceptors handle protocol and model-source quirks such as Anthropic next-gen/legacy models, CPA proxy session affinity, DeepSeek, Responses API adjustments including reasoning effort, built-in image generation tools, and service tier handling, Gemini legacy models, Kimi, MiniMax, OpenCode Go, generic OpenAI-compatible endpoints, OpenRouter metadata and caching, xAI request normalization and native search, Zhipu GLM, and Zenmux model routing without compromising the protocol.
 
 When an image generation model is configured, workflows that allow the `ImageGen` tool can generate images directly from chat and keep the result as an artifact. For direct image and video work, Image Workshop opens with `omx --serve --image-workshop` and supports generation (including video models), edits with reference images and masks, privacy mode, result sessions, and artifact folder access.
 
@@ -78,7 +78,7 @@ omx --remove-provider openrouter
 
 In chat, `/git-commit` can generate a commit message from staged changes with the auxiliary model and run `git commit`.
 
-Built-in providers: **Zenmux**, **Zenmux Expert**, **DeepSeek**, **Kimi for Coding**, **MiniMax Coding Plan**, **OpenCode Go**, **OpenRouter**, **Zhipu Coding Plan**
+Built-in providers: **Zenmux**, **Zenmux Expert**, **DeepSeek**, **Kimi for Coding**, **MiniMax Coding Plan**, **OpenCode Go**, **OpenRouter**, **Synthetic**, **Zhipu Coding Plan**
 
 ## Atlas Navigation
 
@@ -106,7 +106,7 @@ Terminal is home base, but OmniContext CLI extends into every tool you use. The
 - **VS Code Extension** - IDE integration with file context, diagnostics, and diff views
 - **Chrome Extension** - sidebar on any webpage for summarization, data extraction, tab control, screenshots, and browser automation
 - **Office Add-in** - AI panel inside Word, Excel, and PowerPoint
-- **Web Client** - browser UI with login support, session rewind, conversation history, Branch sessions from selected assistant text, LaTeX, Mermaid diagrams, file attachments, drag-and-drop, memory management, tool approval, response/work language settings, per-turn stream metrics (time to first token, total time, tokens per second), and Image Workshop for image and video generation and editing
+- **Web Client** - browser UI with login support, session rewind and conversation forking, conversation history, annotations and Branch sessions from selected assistant text, LaTeX, Mermaid diagrams, file attachments, drag-and-drop, memory management, tool approval, response/work language settings, per-turn stream metrics (time to first token, total time, tokens per second), and Image Workshop for image and video generation and editing
 - **Mobile Access** - run `omx --serve --host 0.0.0.0 --qr` and scan the QR code from your phone; install the web UI to your home screen as a PWA for a standalone app experience
 
 ## Extensibility

package/README.zh-CN.md

@@ -9,7 +9,7 @@
 - **精简系统提示词** - 内置提示词短小、聚焦、没有模板废话。工具描述力求精简。你的上下文预算用在实际工作上，而不是框架开销。
 - **智能体式工具** - 探索性任务（搜索代码、勘察项目结构、预览文件）作为独立的智能体式工具运行，也可以单独指定辅助模型。中间工具调用、文件内容和推理过程不会进入主上下文。你只拿到一个简洁的结论，中间草稿完全不占空间。
 - **代码地图** - 代码导航知识会保存为轻量的主题到文件索引，让助手先跳到关键文件，再决定是否需要继续搜索。
-- **跨会话记忆** - 每次对话压缩时，模型自动提炼关键要点并评估已有记忆。好的见解不断积累权重，过时的知识自然衰减。下次会话开始时，积累的记忆自动注入，让助手从第一句话就了解你的偏好和项目背景。
+- **跨会话记忆** - 每次对话压缩时，模型自动提炼关键要点并评估已有记忆。好的见解不断积累权重，过时的知识自然衰减。新聊天会话开始时，积累的记忆自动注入，让助手从第一句话就了解你的偏好和项目背景。
 - **提示词缓存** - 你可以自主选择缓存 TTL：5 分钟适合快速迭代，1 小时适合长时间工作。Anthropic 的缓存控制自动应用，重复的上下文块直接从缓存读取，大幅降低 token 消耗和成本。
 - **上下文编辑** - 对于很长、工具调用很多的会话，高级设置可以在下一次请求前压缩较早的工具输入、工具输出和推理痕迹，在保留已保存历史的同时减少 token 用量。
 
@@ -61,7 +61,7 @@ allowRemoteTools: true
 | **Google Videos** | 用于图像工坊的 Google 视频生成 |
 | **xAI Videos** | 用于图像工坊的 xAI 视频生成 API |
 
-供应商专属拦截器处理协议和模型源差异，例如 Anthropic 新旧模型、CPA 代理会话亲和、DeepSeek、Responses API 调整与服务层级处理、Gemini 旧模型、Kimi、MiniMax、OpenCode Go、通用 OpenAI 兼容端点、OpenRouter 元数据与缓存、xAI、Zhipu GLM 和 Zenmux 模型路由，同时不影响协议本身。
+供应商专属拦截器处理协议和模型源差异，例如 Anthropic 新旧模型、CPA 代理会话亲和、DeepSeek、Responses API 调整（含推理努力、内置图像生成工具与服务层级处理）、Gemini 旧模型、Kimi、MiniMax、OpenCode Go、通用 OpenAI 兼容端点、OpenRouter 元数据与缓存、xAI 请求规范化与原生搜索、Zhipu GLM 和 Zenmux 模型路由，同时不影响协议本身。
 
 配置图像生成模型后，允许 `ImageGen` 工具的工作流可以直接在聊天中生成图像，并把结果保存为生成物。如果要直接处理图像和视频，可以用 `omx --serve --image-workshop` 打开图像工坊，支持生成（含视频模型）、基于参考图和蒙版的编辑、隐私模式、结果会话和生成物文件夹访问。
 
@@ -78,7 +78,7 @@ omx --remove-provider openrouter
 
 在聊天中，`/git-commit` 可以用辅助模型根据暂存区变更生成提交信息，并运行 `git commit`。
 
-内置供应商：**Zenmux**、**Zenmux Expert**、**DeepSeek**、**Kimi for Coding**、**MiniMax Coding Plan**、**OpenCode Go**、**OpenRouter**、**Zhipu Coding Plan**
+内置供应商：**Zenmux**、**Zenmux Expert**、**DeepSeek**、**Kimi for Coding**、**MiniMax Coding Plan**、**OpenCode Go**、**OpenRouter**、**Synthetic**、**Zhipu Coding Plan**
 
 ## 代码地图
 
@@ -96,7 +96,7 @@ Atlas 是面向大型或高频项目的持久代码导航索引。每张地图
 | 中性 | -1 |
 | 有害 | -6 |
 
-评分低于 -10 的要点被剪枝。好的见解在多次会话中不断积累权重，坏的建议很快被清除，过时的知识自然衰减。每个项目都会把记忆存放在 `~/.omx/projects/<project-id>/memory.json` 里，也可以直接编辑、使用内置记忆工具，或通过 Web 客户端的记忆面板管理。
+评分达到 -10 或更低的要点会被剪枝。好的见解在多次会话中不断积累权重，坏的建议很快被清除，过时的知识自然衰减。每个项目都会把记忆存放在 `~/.omx/projects/<project-id>/memory.json` 里，也可以直接编辑、使用内置记忆工具，或通过 Web 客户端的记忆面板管理。
 
 ## 桌面版与集成
 
@@ -106,7 +106,7 @@ Atlas 是面向大型或高频项目的持久代码导航索引。每张地图
 - **VS Code 扩展** - IDE 集成，感知打开文件、诊断信息和 diff 视图
 - **Chrome 扩展** - 任意网页上的侧边栏，支持摘要、数据提取、标签页控制、截图和浏览器自动化
 - **Office 插件** - Word、Excel 和 PowerPoint 内的 AI 面板
-- **Web 客户端** - 浏览器 UI，支持登录、会话回退、会话历史、从选中的助手文本创建 Branch 会话、LaTeX、Mermaid 图表、文件附件、拖拽、记忆管理、工具审批、回复/工作语言设置、每轮流式指标（首 token 时间、总时长、每秒 token 数），以及用于图像与视频生成和编辑的图像工坊
+- **Web 客户端** - 浏览器 UI，支持登录、会话回退与对话分叉、会话历史、选中助手文本后的标注和 Branch 会话、LaTeX、Mermaid 图表、文件附件、拖拽、记忆管理、工具审批、回复/工作语言设置、每轮流式指标（首 token 时间、总时长、每秒 token 数），以及用于图像与视频生成和编辑的图像工坊
 - **移动端访问** - 运行 `omx --serve --host 0.0.0.0 --qr` 后用手机扫码连接；可将 Web UI 安装到主屏幕作为 PWA，获得独立应用体验
 
 ## 可扩展性

package/dist/assets/cone-mac.icon/icon.json

@@ -0,0 +1,38 @@
+{
+  "fill": {
+    "solid": "srgb:1.00000,1.00000,1.00000,1.00000"
+  },
+  "groups": [
+    {
+      "layers": [
+        {
+          "glass": false,
+          "hidden": false,
+          "image-name": "cone.png",
+          "name": "Cone",
+          "position": {
+            "scale": 1.8,
+            "translation-in-points": [
+              0,
+              0
+            ]
+          }
+        }
+      ],
+      "shadow": {
+        "kind": "none",
+        "opacity": 0.5
+      },
+      "specular": false,
+      "translucency": {
+        "enabled": false,
+        "value": 0.5
+      }
+    }
+  ],
+  "supported-platforms": {
+    "squares": [
+      "macOS"
+    ]
+  }
+}