Skip to content

Axis-Analytica/timbre

Repository files navigation

Timbre

A voice-driven writing tool that learns how you write. Draft LinkedIn posts and Substack articles in your authentic voice, manage a writing pipeline, and generate editorial images — all powered by AI that gets better with your feedback.

Built by Axis Analytica.

How it works

  1. Define your voice — Describe your writing style, tone, and anti-patterns during setup
  2. Write with AI — Enter an idea or paste a brain-dump, and Timbre generates drafts in your voice
  3. Refine over time — Flag what sounds right and what doesn't. Your voice framework sharpens with every piece
  4. Publish anywhere — Preview for LinkedIn or Substack, copy with one click

Features

  • Writing pipeline — Kanban board tracking pieces from Idea to Published
  • Three input modes — Brief (quick idea), Transcription (brain-dump), or direct Edit
  • Voice feedback loop — "Sounds like me" / "Not quite" corrections accumulate and improve generation
  • Platform previews — See how your piece looks on LinkedIn vs Substack before copying
  • Editorial image generation — AI-crafted image prompts in your brand style, with DALL-E integration
  • Multi-provider AI — Works with Claude, GPT-4, or Gemini via the Vercel AI SDK
  • File-based storage — Everything is markdown and JSON. No database. Works with Claude Code remotely
  • Configurable themes — Warm, Cool, or Neutral colour presets, or set your own brand colours

Getting started

git clone https://github.com/Axis-Analytica/timbre.git
cd timbre
npm install
npm run dev

Open http://localhost:3000 — the setup wizard will walk you through configuring your voice, AI provider, and theme.

API keys

You'll need at least one LLM API key. Add them to .env.local:

# Pick your LLM provider
ANTHROPIC_API_KEY=sk-ant-...     # For Claude
OPENAI_API_KEY=sk-...            # For GPT-4
GOOGLE_API_KEY=...               # For Gemini

# Optional — for image generation
IMAGE_API_KEY=sk-...
IMAGE_API_PROVIDER=openai

Project structure

timbre/
├── content/          # Your articles (markdown + YAML frontmatter)
├── voice/
│   ├── style-guide.md      # Your voice framework (editable)
│   ├── examples.md          # Approved writing samples
│   └── corrections.json     # Accumulated voice feedback
├── images/           # Generated image prompts and saved images
├── src/
│   ├── app/          # Next.js App Router pages and API routes
│   ├── components/   # React components
│   └── lib/          # Data layer, AI client, config
└── timbre.config.json  # Your personal config (created during setup)

Views

View Path Purpose
Pipeline / Kanban board — drag pieces between Idea, Draft, Review, Published
All Writing /writing Searchable list, filterable by platform
Writing Workspace /write/[slug] Editor with AI generation, platform preview, voice feedback
Voice Guide /voice Edit your voice framework, view correction history
Image Studio /images Generate editorial image prompts, preview and save
Settings /settings API key reference

Tech stack

  • Next.js 16 (App Router)
  • React 19, TypeScript, Tailwind CSS 4
  • Vercel AI SDK (multi-provider: Anthropic, OpenAI, Google)
  • OpenAI SDK (DALL-E image generation)
  • gray-matter (frontmatter parsing)
  • File-based storage (markdown + JSON)

License

MIT

About

A voice-driven writing tool that learns how you write

Resources

License

Stars

1 star

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors