I got tired of dictation apps mangling my variable names, so I built one that reads the repo first

reddit SideProject · 原文链接 ↗ · 63 分

I talk to Claude Code and Cursor most of the day, and every dictation app I tried broke on the same thing. I'd say "useAuthSession" and get "use auth session". "kCGEventTap" came out as "KCG event tap". Fixing the names afterwards took longer than just typing them, so I kept going back to the keyboard. The fix turned out to be simpler than I expected: look at the repo before you listen. It pulls the symbols out of whatever project has focus and biases decoding toward them, so your actual identifiers beat the phonetically similar English. Switch projects and it re-indexes on focus change,

五维评分

AI 分析

痛点：通用听写应用无法准确识别代码变量名（如将 useAuthSession 识别为 use auth session），导致开发者修正时间比打字还长，影响编码效率。

目标用户：使用语音听写辅助编程的开发者，尤其是频繁与 Claude Code、Cursor 等 AI 编程助手对话的人群。

证据：I'd say "useAuthSession" and get "use auth session". "kCGEventTap" came out as "KCG event tap". Fixing the names afterwards took longer than just typing them, so I kept going back to the keyboard. ... Wispr Flow sends your audio to their servers. Superwhisper and MacWhisper can run locally, but they use general-purpose models, so they hit the same identifier problem I started with.

变现：订阅制或一次性买断，类似 Superwhisper 和 Wispr Flow 的定价模式；也可作为现有听写工具的付费插件。

MVP：用本地 Whisper 或 Apple Speech 框架构建 macOS 听写工具，增加代码仓库索引步骤：解析当前项目文件提取标识符，在解码时偏向这些词。2周内可完成基础版，实现低延迟与本地运行。

竞品：Wispr Flow、Superwhisper、MacWhisper

摘要：一位开发者因通用听写应用无法正确识别代码变量名（如将 useAuthSession 变成 use auth session）而自建了能读取代码仓库的听写应用。该应用在本地运行，延迟190ms，代码词准确率97.2%，已有Wispr Flow等付费竞品。这揭示了开发者听写场景下的细分需求。