ai-agent

fcitx5-vinput

Name: fcitx5-vinput
Author: xifan2333

Voice input for Fcitx5 — local and cloud ASR, LLM rewriting, cross-distro packages

599 stars+39★ this week52 forksC++Updated 10/23/2026100% free · open source

What it does

Adds voice‑to‑text input to Fcitx5, letting you dictate locally (Vosk) or via cloud (OpenAI) and optionally rewrite the text with an LLM.

When to use it

•You want hands‑free typing on Linux desktops that already use Fcitx5.
•You need a privacy‑focused local speech recogniser (Vosk) but also want the option to call a cloud LLM for rewriting.
•You prefer a single install that works across Ubuntu, Arch, and Fedora without pulling in heavyweight GUI dictation apps.

Ready-to-paste prompt

Press the hotkey, say: “Schedule a meeting with the product team next Tuesday at 10 am, then rewrite it as a formal email invitation.” The plugin will transcribe, send the raw text to OpenAI, and paste the polished email into your editor.

Heads up: Cloud transcription and LLM rewriting require a valid OpenAI API key; without it the plugin falls back to Vosk, which only supports the languages for which you have downloaded a model.

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/fcitx5-vinput && curl -fsSL https://workflowstacks.com/api/skills/fcitx5-vinput/claude-skill -o ~/.claude/skills/fcitx5-vinput/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How fcitx5-vinput works

Codeflow

Free to inspect

Fcitx5-vinput is a very large C++ project (~79k lines across 352 code files, plus 32 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, GPL-3.0 license, has a test suite.

Size

Very large codebase

~79k lines · 352 code files · days to read — use, don't read

Setup

Developer setup

A real software project. Use it via its install path; don't expect to read it all.

Runs on

Python

No API keys detected

C++ 89%Roff 3%CMake 2%Shell 2%MDX 1%

Where to start reading

1
README.md
Start here — what it does and how to install it
2
AGENTS.md
The instructions the AI actually follows
3
pyproject.toml
Dependencies and the commands it exposes
4
src/cli/main.cpp
Inside src/ — the main logic begins here

What's in each folder

src/Core code — the actual logic209 files

site/Frontend / UI31 files

3rdparty/Folder152 files

data/Data files39 files

packaging/Folder16 files

scripts/Helper scripts9 files

po/Folder2 files

systemd/Folder2 files

READMEHas testsDocumentedCI checksGPL-3.0 licenseUpdated this month

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/fcitx5-vinput && curl -fsSL https://workflowstacks.com/api/skills/fcitx5-vinput/claude-skill -o ~/.claude/skills/fcitx5-vinput/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

Quick Actions

Details

Creator

xifan2333

Language

C++