How lfnovo/ai-prompter works
AI Prompter is a small Python library for finding or supplying Jinja templates, rendering prompts from data, and optionally exporting them to LangChain. The main workflow is template setup followed by rendering; templates can be composed with includes. The README also documents structured-output parser support. The only runtime source file is src/ai_prompter/__init__.py; some documented features are not visible in the sampled excerpts, so their internal wiring is not inferred.
Main entry points: __init__.py Prompter
Components 11
Other components
- Application developer
- Prompter
src/ai_prompter/__init__.py - Search paths
src/ai_prompter/__init__.py - Template directories
- Jinja loader
src/ai_prompter/__init__.py - Jinja template
src/ai_prompter/__init__.py - Render data
- Rendered prompt
- LangChain export
src/ai_prompter/__init__.py - LangChain
- Output parser support
src/ai_prompter/__init__.py
Connections 14
- Application developer → Prompter: creates
- Prompter → Search paths: builds paths
- Search paths → Template directories: selects
- Prompter → Jinja loader: sets up
- Jinja loader → Template directories: loads from
- Jinja loader → Jinja template: loads template
- Application developer → Prompter: provides text
- Prompter → Jinja template: creates from text
- Application developer → Prompter: provides data
- Prompter → Jinja template: renders
- Jinja template → Rendered prompt: produces
- Prompter → LangChain export: requests conversion
- LangChain export → LangChain: creates prompt
- Prompter → Output parser support: supports