# Promptel

> Declarative prompt engineering. A small DSL (and equivalent YAML) for LLM prompts, with typed params, technique blocks, and a provider abstraction over OpenAI / Anthropic / Groq.

Project: Skelf Research / promptel
License: MIT
Source: https://github.com/skelf-research/promptel
Docs: https://docs.skelfresearch.com/promptel/

## Pages
- /                         — overview, features, format comparison
- /about/                   — what it is, what it isn't, scope
- /blog/                    — notes index
- /blog/why-prompts-deserve-a-dsl/
- /blog/provider-portable-prompts/
- /blog/anatomy-of-a-promptel-file/
- /compare/                 — comparisons index
- /compare/prompty/         — vs Microsoft Prompty
- /compare/dspy/            — vs Stanford DSPy
- /rss.xml                  — feed
- /llms-full.txt            — fuller machine-readable summary

## Summary
- Authoring: `.prompt` DSL parsed with Chevrotain, or equivalent YAML
- Runtime: Node.js ≥ 18, ships as `promptel` on npm
- Providers: OpenAI, Anthropic Claude, Groq (official SDKs)
- Techniques: Chain-of-Thought, Few-Shot, Zero-Shot, Tree-of-Thoughts, ReAct, Self-Consistency
- Harmony Protocol: native multi-channel response support
- CLI: execute and convert prompts (`promptel`)