# microsoft/promptbase

All things prompt engineering

Repository: https://github.com/microsoft/promptbase
Canonical: https://ross.abutalabs.com/products/promptbase
Language: Python
License: MIT
License Family: permissive
Last push: 2024-06-04T13:26:33+00:00

## Health v2 (maintenance only)
Score: 26/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 71
- inputs: {"age_days": 995, "days_push": 820, "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 5771, forks 332 (observed 2026-08-28T04:09:29.443823+00:00)

## What it is
A Microsoft-maintained collection of prompt engineering resources, best practices, and example Python scripts for eliciting top performance from foundation models like GPT-4. It hosts the Medprompt methodology (dynamic few-shot selection, self-generated chain of thought, and choice-shuffle ensembling) with benchmark results across MMLU, GSM8K, and other evaluations.

## Use cases
- improve gpt-4 accuracy with prompting techniques
- learn medprompt methodology
- benchmark llm prompting strategies
- apply chain-of-thought and few-shot prompting
- replicate medprompt+ benchmark results
- study prompt engineering best practices

## When to choose
- you want proven prompting techniques like Medprompt with benchmark-backed results
- you need example scripts for few-shot selection and chain-of-thought ensembling
- you are researching how prompting compares to fine-tuning

## When to avoid
- you need a production prompt management or deployment tool
- you want a maintained library with an API rather than example scripts
- you work with small local models where these techniques may not transfer

## Facets
- artifact type: learning-resource
- maturity: active
- function: prompt-engineering, llm-inference, benchmarking
- domain: large-language-models, artificial-intelligence, tutorials
- platform: python
- tags: medprompt, few-shot-learning, chain-of-thought, gpt-4, research

## Member repositories
- microsoft/promptbase (main) score 26

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:29.443823+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 2026-08-29T17:52:54.722606+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/promptbase (fetched 2026-08-28T04:09:29.443823+00:00, sha 43c738286a74)
- Data as of 2026-08-30T08:39:29.467469+00:00.