How vaishnavisukum/quilly works
Quilly turns a YouTube or plain-text transcript into normalized transcript JSON, then generates LinkedIn carousels, a newsletter, and FAQs as publishing-ready files. Users can invoke each step through the CLI or describe an action in natural language; the latter is routed through Gemini and dispatched by the router. Transcript handling also loads and caps inputs, while shared prompt/context and Gemini-client layers support routing and generation. Relationships are limited to calls and data flows shown in excerpts or README; output-file nodes represent artifacts, not inspected source modules.
Main entry points: router.py Command router · generate.py Generation orchestration · cli.py CLI commands · client.py Gemini client
Components 18
CLI and routing
- CLI commands
src/quilly/cli.py - Command router
src/quilly/router.py
Transcript handling
- YouTube fetch
src/quilly/transcript.py - Text conversion
src/quilly/transcript.py - Load and cap
src/quilly/transcript.py
Content generation
- Router prompt
src/quilly/prompts.py - Generation orchestration
src/quilly/generate.py - Content prompts and context
src/quilly/prompts.py - Carousel generation
src/quilly/generate.py - Newsletter generation
src/quilly/generate.py - FAQ generation
src/quilly/generate.py
External services
- Gemini client
src/quilly/client.py - Gemini API
- YouTube captions
Other components
- User
- Plain-text transcript
- Transcript JSON
- JSON and Markdown assets
Connections 29
- User → CLI commands: runs command
README.md - CLI commands → YouTube fetch: calls fetch
src/quilly/cli.py - CLI commands → Text conversion: calls conversion
src/quilly/cli.py - CLI commands → Command router: routes ask
src/quilly/cli.py - CLI commands → Generation orchestration: calls generation
src/quilly/cli.py - Command router → Router prompt: uses prompt
src/quilly/router.py - Command router → Gemini client: requests routing
src/quilly/router.py - Gemini client → Gemini API: calls model
src/quilly/client.py - Command router → YouTube fetch: dispatches fetch
src/quilly/router.py - Command router → Text conversion: dispatches conversion
src/quilly/router.py - Command router → Generation orchestration: dispatches generation
src/quilly/router.py - YouTube fetch → YouTube captions: fetches captions
src/quilly/transcript.py - Plain-text transcript → Text conversion: provides text
README.md - YouTube fetch → Transcript JSON: writes entries
src/quilly/transcript.py - Text conversion → Transcript JSON: writes entries
src/quilly/transcript.py - Generation orchestration → Load and cap: loads and caps
src/quilly/generate.py - Load and cap → Transcript JSON: reads JSON
src/quilly/transcript.py - Generation orchestration → Carousel generation: generates type
src/quilly/generate.py - Generation orchestration → Newsletter generation: generates type
src/quilly/generate.py - Generation orchestration → FAQ generation: generates type
src/quilly/generate.py - Carousel generation → Content prompts and context: uses prompt and context
src/quilly/generate.py - Newsletter generation → Content prompts and context: uses prompt and context
src/quilly/generate.py - FAQ generation → Content prompts and context: uses prompt and context
src/quilly/generate.py - Carousel generation → Gemini client: requests generation
src/quilly/generate.py - Newsletter generation → Gemini client: requests generation
src/quilly/generate.py - FAQ generation → Gemini client: requests generation
src/quilly/generate.py - Carousel generation → JSON and Markdown assets: writes JSON and Markdown
src/quilly/generate.py - Newsletter generation → JSON and Markdown assets: writes Markdown
src/quilly/generate.py - FAQ generation → JSON and Markdown assets: writes JSON and Markdown
src/quilly/generate.py