# microsoft/prompty

Prompty makes it easy to create, manage, debug, and evaluate LLM prompts for your AI applications.  Prompty is an asset class and format for LLM prompts designed to enhance observability, understandability, and portability for developers.

Repository: https://github.com/microsoft/prompty
Canonical: https://ross.abutalabs.com/products/prompty
Homepage: https://prompty.ai
Language: Rust
License: MIT
License Family: permissive
Topics: generative-ai, llm-evaluation, llms, promptengineering, prompty
Last push: 2026-08-25T05:01:21+00:00

## Health v2 (maintenance only)
Score: 71/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 40, longevity 61
- inputs: {"age_days": 863, "days_push": 8, "days_rel": 405, "gap_med": 0.0, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1252, forks 126 (observed 2026-08-28T04:04:08.449677+00:00)

## What it is
Prompty is a markdown-based file format (.prompty) for authoring LLM prompts with YAML frontmatter for model config, inputs, tools, and templates. It ships runtimes for Python, TypeScript, and Rust plus a VS Code extension to execute, trace, and debug prompts across providers like OpenAI, Anthropic, and Microsoft Foundry.

## Use cases
- manage llm prompts as versioned files
- run a prompt from python or typescript without rewriting it
- debug and trace llm prompt executions
- evaluate and iterate on prompt templates
- switch llm providers without changing prompt files
- author prompts in vs code with syntax highlighting
- render jinja2 or mustache prompt templates

## When to choose
- you want a portable, single-file prompt asset format across languages
- you need built-in tracing and observability for prompt runs
- your team authors prompts in VS Code and executes them in Python or TypeScript
- you need to swap LLM providers or endpoints without editing prompt bodies

## When to avoid
- you need a full agent orchestration framework rather than prompt assets
- you require a stable, frozen API - the v2 format and tooling are in alpha and may change
- you only call LLMs with simple hardcoded strings and don't need templating or tracing

## Facets
- artifact type: library
- maturity: experimental
- function: prompt-engineering, llm-inference, templating, tracing, sdk
- domain: large-language-models, developer-tools, artificial-intelligence
- platform: python, rust, cross-platform, editor-plugin
- tags: prompt-format, markdown-prompts, vscode-extension, llm-observability, portable-prompts, jinja2, mustache, ai-agents, nodejs

## Member repositories
- microsoft/prompty (main) score 71

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:08.449677+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T05:07:33.975531+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/prompty (fetched 2026-08-28T04:04:08.449677+00:00, sha 96c7c8e6a208)
  - homepage: https://prompty.ai (fetched 2026-08-29T12:18:15.579446+00:00, sha 4b483a64c024)
  - registry_crates: https://crates.io/api/v1/crates/prompty (fetched 2026-08-29T12:18:15.588414+00:00, sha 923f3a2341b8)
- Data as of 2026-08-30T08:39:29.467469+00:00.