Find software, follow a workflow, and keep the instructions that make it work.
Tools may require your own subscription, API account, or local setup.
Choose a task. See what to bring and what to check.
Research workflowUse an AI product description workflow to research buyer questions, check product facts, draft useful copy, and review it against original sources.
Data workflowRun product spreadsheet cleanup with OpenRefine: trim names, standardize category casing, preserve IDs, and review exceptions before reuse.
Presentations workflowUse an AI mind map workflow to organize project notes into a checked outline, keep tentative ideas separate, and render it with Markmap.
Automation workflowPrepare an AI weekly report from complete source windows, checked totals, explicit unknowns, and a practical next action.
Writing workflowPrepare AI meeting minutes from a checked transcript, preserving action owners, conditional deadlines, unclear passages, and team confirmation.
Research workflowBuild a supplier quote comparison with source-linked quantities, currencies, minimum orders, shipping, and unresolved terms before choosing.
Know where a tool fits before adding it to your setup.
Use OpenRefine to clean supplier CSV names and categories, preserve product IDs, review exceptions, and save repeatable operations.
Explore guide ↗Set up GPT Researcher with your own service accounts, run a bounded question, and check the original sources before using its report.
Explore guide ↗Set up Vale for editorial style checks with an original rule, a sample document, and a separate factual review.
Explore guide ↗Use Markmap to render a checked Markdown mind map, review its branches, and keep the source outline for later edits.
Explore guide ↗Use Grist to import a practice CSV, check dates and data types, and reconcile totals before preparing a report.
Explore guide ↗Build a small Activepieces flow with sample inputs, validation, a draft output, and an explicit review before external actions.
Explore guide ↗Use whisper.cpp to transcribe a permitted recording locally, export text and SRT, and check names, numbers, and conditions against the audio.
Explore guide ↗Use Docling in your own Python environment to extract practice PDFs, check source fields, and prepare a supplier comparison.
Explore guide ↗Better material, clearer instructions, and a checkable result.
Build an AI context pack with the reader decision, original evidence, unresolved facts, constraints, and a definition of usable output.
Read the guide ↗Use an AI output review to check claims and conditions against original sources, separate factual corrections from style, and record open questions.
Read the guide ↗Save a reusable AI workflow kit with required inputs, procedure, original examples, review checks, and recovery notes.
Read the guide ↗Write AI task instructions with defined inputs, ordered steps, a checkable deliverable, stop conditions, and a reviewer handoff.
Read the guide ↗Reusable material you can save and adapt.
A place to record claims, source links, limitations, and open questions before writing.
Preview the template ↗Check evidence, numbers, instructions, and the next action before using a result.
Preview the template ↗Three prompts, an evidence ledger, a fictional worked example, and a review checklist for a product-page draft.
Preview the template ↗Practice a repeatable supplier-CSV cleanup with original data, an expected result, two expressions, and a review checklist.
Preview the template ↗Organize meeting notes with two prompts, a source-linked example outline, and a team-review checklist. Includes three Markdown downloads.
Preview the template ↗Practice a checked weekly update with dated CSV inputs, reference totals, an AI drafting prompt, and a review checklist.
Preview the template ↗Choose a meeting-notes route, practice with a written transcript, and prepare checked actions and a team confirmation message.
Preview the template ↗Choose manual entry or extraction, compare practice quotes, and use a blank buying brief to prepare supplier questions.
Preview the template ↗A useful recommendation needs context. Look for the task, setup, running costs, and checks alongside each tool.
Our approach ↗