# microsoft/PromptWizard

Task-Aware Agent-driven Prompt Optimization Framework

Repository: https://github.com/microsoft/PromptWizard
Canonical: https://ross.abutalabs.com/products/promptwizard
Language: Python
License: MIT
License Family: permissive
Last push: 2025-10-13T21:51:01+00:00

## Health v2 (maintenance only)
Score: 45/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 46, release rhythm 35, longevity 58
- inputs: {"age_days": 825, "days_push": 324, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4007, forks 353 (observed 2026-08-28T04:08:32.200287+00:00)

## What it is
PromptWizard is a discrete prompt optimization framework from Microsoft that uses a self-evolving mechanism where an LLM generates, critiques, and refines its own prompts and in-context examples. It iteratively optimizes instructions and examples together, including synthetic examples and self-generated chain-of-thought reasoning steps.

## Use cases
- optimize prompts for better LLM task performance
- automatically generate synthetic few-shot examples
- improve prompt accuracy with chain-of-thought reasoning
- tune system prompts without manual trial and error
- optimize prompts using my own training data
- evolve instructions and in-context examples together

## When to choose
- you want automated, feedback-driven prompt optimization instead of manual iteration
- you need to jointly optimize prompts and few-shot examples
- you want a research-backed framework with MIT licensing

## When to avoid
- you need a hosted no-code prompt tuning service
- your use case involves fine-tuning model weights rather than prompts
- you cannot spend LLM API tokens on iterative critique and refinement

## Facets
- artifact type: library
- maturity: active
- function: prompt-engineering, llm-inference, machine-learning
- domain: large-language-models, artificial-intelligence, developer-tools
- platform: python, cross-platform
- tags: prompt-optimization, chain-of-thought, synthetic-data-generation, in-context-learning, self-evolving-prompts, natural-language-processing

## Member repositories
- microsoft/PromptWizard (main) score 45

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:32.200287+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 2026-08-29T18:24:03.458738+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/PromptWizard (fetched 2026-08-28T04:08:32.200287+00:00, sha 2197a5a89fa4)
- Data as of 2026-08-30T08:39:29.467469+00:00.