Prompt library

Worked starting points, each already matched to a framework and a model. Open one in the builder and edit it into your own situation — the examples are deliberately specific so you can see what a filled-in slot should look like.

Business & management

GCSEMicrosoft

Status update for a steering group

Pulls a monthly project update out of documents and chat you already have.

GCSEMicrosoft

Meeting notes into decisions and actions

Separates what was decided from what was merely discussed.

TAGAnthropic

Retrospective into a roadmap shortlist

Three sentences, where the Goal slot does the real steering.

Marketing & communication

CO-STAROpenAI

Customer email with a specific voice

Uses the three voice slots to keep marketing copy off the default register.

APEAnthropic

Critique of page copy

Asks for findings tied to specific sentences, so you can act on them.

Engineering

CRAFTAnthropic

Focused code review

Bounds the review so you get real findings instead of a style lecture.

RTFOpenAI

Error messages into documentation

A three-slot prompt where the format instruction does most of the work.

HR & policy

RISENAnthropic

Internal policy draft

The Nuance slot carries the political constraints that sink most first drafts.

RACEOpenAI

Interview questions that discriminate

Context carries what went wrong last time, which is what stops a repeat.

Education & research

TCREIGoogle

Lesson plan with worked examples

The References slot is what makes the output match your existing materials.

CRAFTGoogle

Summary that keeps the caveats

Explicitly asks for limitations, which summaries otherwise drop first.

Data

RTFOpenAI

Structured extraction to JSON

Names the keys and forbids the preamble — the two things that break parsers.