Repository navigation
Herta 0.1.10
黑塔 v0.1.10 — 内置浏览器、信息流、引用 Word
本次更新
内置浏览器 查看器新增网页标签页。工作区 herta.launch.json 中的开发服务器可在新标签页中启动；差分协处理器可以打开本机页面并阅读、点击、输入与截图，每一步都会记入对话记录。其操作仅限本机地址，启动开发服务器的命令按权限规则确认。
信息流 按「设置 → 信息流」中选择的话题，每天整理一期网上的新内容，附摘要、正文与黑塔的短评，可在内置浏览器中查看原文，或引用到对话。默认关闭，开启后每期将额外消耗 DeepSeek 额度。
引用 Word 文档 Word 文档按章节分块，可像 PDF 一样分段引用，点击引用标签可定位到原段落。
安全与稳定性 修复了 macOS 上命令权限检查的若干漏洞，经 caffeinate、nice 等包装的危险命令现在会被拦截；以明文保存的密钥会改为加密保存。此外修复了引用、查看器与界面方面的若干问题。
Downloads
| 平台 | 文件 |
|---|---|
| Windows | Herta-Setup-0.1.10.exe |
| macOS (Apple silicon) | Herta-0.1.10-arm64.dmg |
| macOS (Intel) | Herta-0.1.10-x64.dmg |
| Linux (x64) | Herta-x86_64.AppImage |
macOS 构建已签名并通过 Apple 公证。已安装 0.1.9 的 Windows、macOS 与 Linux（AppImage）用户将自动收到更新。
Linux 版下载后需先执行 chmod +x Herta-x86_64.AppImage 再运行，之后的版本可在应用内更新。在限制非特权用户命名空间的发行版上（如 Ubuntu 24.04 的默认配置），AppImage 会在关闭 Chromium 沙箱的情况下运行，此时内置浏览器宜仅用于本机开发服务器。
使用前需配置 DeepSeek API 密钥，密钥仅保存在本机；自动审核、输入预测、PDF 图片转写、网页搜索与信息流也使用该密钥。语音模型在首次启用实时语音时下载。
第三方声明 安装包内的语音合成运行时是 sherpa-onnx 1.13.6 的预编译包，其中静态链接了 espeak-ng（GPL-3.0-or-later）与 piper-phonemize（MIT）。对应源码：k2-fsa/sherpa-onnx v1.13.6 的构建树，以及它所拉取的 csukuangfj/espeak-ng@ed530aa 与 csukuangfj/piper-phonemize@f3ff95a。完整的许可证文本随应用安装在资源目录的 THIRD-PARTY-NOTICES.md 中。
In English
Built-in browser — The file viewer gains a page tab. The dev servers listed in the workspace's herta.launch.json start from the new-tab page; the Coprocessor can open pages on this machine and read, click, type and take screenshots, and each step is recorded in the conversation. Its steps stay on local addresses, and starting a dev server is confirmed by the permission rules.
Flow — One issue a day of new items from the web on the topics you pick in Settings → Flow, each with a summary, a write-up and a short word from Herta; the source opens in the built-in browser, and an item can be quoted into the conversation. Off by default; once on, each issue uses additional DeepSeek credit.
Quoting Word documents — Word documents split into sections and can be quoted piece by piece, as PDFs can; clicking a quote's label opens the document at that paragraph.
Safety and stability — Several gaps in the command permission checks on macOS are closed: dangerous commands wrapped in caffeinate, nice and the like are now blocked, and a key stored in plain text is moved to encrypted storage. Several issues with quotes, the viewer and the interface are fixed.
macOS builds are signed and notarized by Apple. Existing 0.1.9 installs on Windows, macOS and Linux (AppImage) update automatically.
On Linux, run chmod +x Herta-x86_64.AppImage once before starting it; later versions update in the app. On distributions that restrict unprivileged user namespaces (Ubuntu 24.04's default, for example), the AppImage runs with Chromium's sandbox turned off, and the built-in browser is then best kept to your own dev servers.
A DeepSeek API key is required; it is stored only on your machine, and auto-review, composer predictions, PDF picture transcription, web search and Flow use it too. The voice model downloads the first time real-time voice is enabled.
Third-party notices — the speech-synthesis runtime in the installer is sherpa-onnx 1.13.6's prebuilt package, which statically links espeak-ng (GPL-3.0-or-later) and piper-phonemize (MIT). Corresponding source: the k2-fsa/sherpa-onnx v1.13.6 build tree and the csukuangfj/espeak-ng@ed530aa and csukuangfj/piper-phonemize@f3ff95a commits it fetches. The full license texts are installed with the app in THIRD-PARTY-NOTICES.md under its resources directory.