<div align="center">

# Promptfire

[![Obsidian](https://img.shields.io/badge/Obsidian-Plugin-7C3AED?logo=obsidian&logoColor=white)](https://obsidian.md)
[![TypeScript](https://img.shields.io/badge/TypeScript-5.8-3178C6?logo=typescript&logoColor=white)](https://www.typescriptlang.org/)
[![License](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
[![Version](https://img.shields.io/badge/Version-1.0.0-blue.svg)](package.json)

**Copy your vault's conventions, structure, and content to any LLM with a single hotkey.**

</div>

## The Problem

When working with AI assistants on your Obsidian vault, you constantly re-explain your folder structure, link style, tag conventions, and frontmatter requirements. This is repetitive, error-prone, and wastes tokens.

## The Solution

**Promptfire** stores your vault's conventions in a dedicated folder and copies everything to clipboard with one hotkey. Output adapts automatically for different LLMs (Claude, GPT-4, Gemini, etc.), and a built-in history tracks every context you generate.

## Key Features

- **Multi-LLM output targets** with per-model token limits, formats (XML, Markdown, Plain), and truncation strategies
- **Prompt templates** with placeholders and conditionals for reusable workflows
- **Additional context sources** from freetext, external files, or shell commands
- **Frontmatter presets** to configure context per-note via `ai-context` YAML
- **Granular section selection** to include only the headings you need
- **Context history** with diff, search, and one-click restore
- **Context generator** that walks you through setting up vault conventions

See the [full documentation](docs/) for details on each feature.

## Quick Start

**1. Generate context files**
```
Ctrl+P > "Promptfire: Generate context files"
```

**2. Add output targets**
Settings > Promptfire > Output Targets > "Add Built-in Targets"

**3. Copy & paste**
```
Ctrl+P > "Promptfire: Copy context to clipboard"
```
Paste into Claude, ChatGPT, or any AI assistant.

## Documentation

| Topic | Description |
|-------|-------------|
| [Getting Started](docs/getting-started.md) | Installation, first run, commands |
| [Output Targets](docs/output-targets.md) | Multi-LLM formats and truncation |
| [Templates](docs/templates.md) | Prompt templates, placeholders, conditionals |
| [Context Sources](docs/context-sources.md) | Freetext, file, and shell sources |
| [Frontmatter Presets](docs/frontmatter-presets.md) | Per-note ai-context YAML configuration |
| [Section Selection](docs/section-selection.md) | Granular heading and block selection |
| [History](docs/history.md) | Context history, diff, and restore |
| [Settings](docs/settings.md) | Full settings reference |
| [Advanced](docs/advanced.md) | Truncation strategies, priority order, XML format |

## Contributing

Contributions welcome! Fork the repo, create a feature branch, run `npm run build` to verify, and submit a pull request. Report bugs and request features via [Codeberg Issues](https://codeberg.org/luca-tty/obsidian-promptfire/issues).

## License

MIT License - see [LICENSE](LICENSE) for details.