← Back to Portfolio

Cognitive Science • AI Optimisation

Prompt Optimiser

A higher-leverage prompt enhancer: structured prompts, instruction hierarchy, output contracts, variables, few-shot examples, clarifying questions, prompt linting, exports (Markdown/JSON), and version history — while keeping the same visual system.

Input Prompt

Intent & Output Contract

Output contract = explicit deliverables + format + constraints. This reduces ambiguity and increases reliability.

Advanced controls constraints, examples, variables, evaluation
Guardrails

These guardrails are phrased to improve reliability without demanding internal chain-of-thought.

Few-shot examples
0 examples added.
Variables
0 variables added. Use placeholders like {{topic}}
Evaluation

A rubric helps the model self-verify before answering (without exposing hidden reasoning).

Medium (3)
Medium (3)
Medium (3)

Output

~0 tokens compression: off
Prompt Analysis
Word Count
0
Readability
0
Specificity
0
Structure
0
Tokens (est.)
0
Ambiguity
0
Optimisation Tips (now dynamic)
Specify deliverables and an output format (the “output contract”).
Add constraints (what to avoid) and success criteria (what “good” looks like).
Use examples for style/format (few-shot) when results are inconsistent.
Ask clarifying questions when requirements are underspecified.