Promptbook — Prompt Library, Variables & Version Tests

Promptbook — Prompt Library, Variables & Version Tests — a premium, 100% offline digital tool by Aceplore. Runs entirely in your browser: no accounts, no subscriptions, your data never leaves your device.

A private, offline library for the prompts you write yourself — every version kept, every {{variable}} declared and checked in both directions, a real line-by-line diff, and a success rate that comes only from trials you recorded — 32 views in one, and it says so plainly when the sample is too small to call a winner.

32 built-in tools in one app. A prompt library with folders, tags and a search that looks inside bodies, version notes, variables, snippets and trial notes; a full version history per prompt with a version-detail page, a line-by-line diff, a side-by-side compare carrying each version’s recorded results, and a rollback that copies an older body forward as a new version with your reason attached — nothing is deleted and the version you left keeps its trials; variables declared per prompt, a fill screen that assembles the text and flags any value you left blank instead of leaving a placeholder in, a missing-variables audit that reports undeclared placeholders AND declarations never used, and a preview of the final substituted text; a snippet store with a compose screen that stacks snippets in order, an editable structure convention of role, task, constraints, format and examples, a lint pass over your own wording, and your own checklists; trials you record by hand, results as succeeded divided by ran, a winner call that will not name a winner off a sample below your minimum, a sample-size rule of thumb with the working shown, and a length view where characters, words and lines are exact counts and the token figure is an estimate from a ratio you set; plus usage measured only by the trials you logged, an activity timeline, plain-text reports, a method page stating every formula it uses, CSV and text export, a JSON backup and settings.

Diff. A longest-common-subsequence comparison over lines, not a character count. Edit one line of a three-line prompt and it reports 1 added, 1 removed and 2 unchanged, so the rule you added is the thing you see.

Missing variables. Given “Act as {{role}}. Summarise {{topic}} for {{role}} in {{tone}}.” with role, tone and audience declared, it finds 3 distinct placeholders — role counted twice but listed once — flags topic as used and never declared, and audience as declared and never used.

Winner. 21 of 30 is 70%, 11 of 20 is 55%, a 15-percentage-point gap — called only because both versions clear the minimum. Drop one to 19 trials and the same lead is too thin to call; raise the minimum to 30 and 20 trials is thin too. A 4-percentage-point gap is too close to call, equal rates are a tie, and a version with no trials has no rate rather than 0%.

What's included:

• Promptbook.html
• Quick-Start-Guide.html
• How-To-Use-Guide.html
• README.txt
• LICENSE.txt

Includes a personalized welcome wizard, an interactive guided tour, a ⌘K command palette, light/dark themes with 8 accents, and one-click JSON backup & restore.

Promptbook is a planning, costing and record-keeping tool for people building with AI. Every figure it reports is computed by a stated formula from the rates, sizes, counts and thresholds YOU enter. It ships no model prices, no context windows and no tokeniser: those differ by provider and model and change without notice, so you enter them and the app shows which ones it used. Any token figure derived from the length of a text is an ESTIMATE, not a count — real tokenisers are model-specific and cannot run offline — so treat it as a planning figure and confirm actual usage against your provider's own reporting and billing before you commit to a budget. Nothing here is a guarantee of a cost, a saving, a quality level or a result, and none of it is business, financial, legal or engineering advice. You are solely responsible for your own decisions.