Produce four pieces on a weekly system with hooks, structure, honest titles, repurposing and a monthly review.
Plan four pieces with three hook options each, timed outlines and reuse plans. Record and edit each on the fixed weekly schedule. Publish with titles under sixty characters and thumbnails under four words. Produce three clips and a newsletter section per piece. Review weekly and write a one-page monthly review with retention by piece and one change.
Model, texture, light and render one product in three views, and export it for the web.
Choose a real product with a reference. Model it at scale with clean topology, unwrap it, build its materials, light it with a studio setup, render three views (hero, detail, in-context), and export a glTF that opens correctly in a web viewer.
Build a themed spawn area, a resource pack that supports its theme, and a plugin that welcomes players and points them to the rules — packaged for a server to install.
Plan and build a spawn with a clear silhouette, palette and landscaping; create a resource pack that changes three textures and a sound to match; write a plugin with a /rules command, a permission, configurable messages and a join greeting with persistent join counts. Package all three with READMEs and versions.
Build a complete shop resource: configurable locations and items, a NUI menu, server-validated purchases through the framework, and a documented release.
Build a shop resource with locations and items in a shared config, a marker and prompt near each location, a NUI menu, a server purchase handler that validates distance, price and balance through one framework wrapper, rate limiting and logging on the handler, and a README, config, version and changelog. Test on staging with a modified client attempting bad purchases.
Rewrite ten listings, fix delivery and policies, build support snippets and macros, run one review cycle and one retention action — and measure.
For a real or realistic store: rewrite the ten best-selling listings to the buyer's-questions structure with spec blocks, set delivery to a library with logged downloads and quotable policies at checkout, build ten snippets and five macros, run one monthly review clustering with listing fixes shipped, and launch one retention email measured by cohort.
Position a product, commit to two channels, run one campaign and an email programme for a quarter, and produce three monthly reviews.
For a real product: write the positioning statement from customer language with proof, choose two channels on evidence with a metric each, launch the email programme with a promise and welcome sequence, run one campaign from a one-page brief with tagged links, and produce three monthly reviews with fair comparisons and caveats.
Take a real business question end to end: definition, profile, query with reconciliation, chart, one-page write-up, and a review by a sceptic.
Choose a question someone in your business actually asks. Write the metric definition and get it agreed. Profile the data and write notes. Build the query with a reconciliation check. Produce one chart with the finding as its title and a caveat beneath. Write the one-page analysis and have a sceptical colleague review it; record their questions and revise.
Build an inventory workbook with a products table, stock movements, reorder points, a dashboard and checks.
Build a workbook with a products table (SKU, name, cost, reorder point, lead time), a movements table (date, SKU, quantity in or out), a stock-on-hand summary with SUMIFS, reorder alerts with conditional formatting, a pivot dashboard by category, validation on all inputs and a checks block.
Implement the money, customer, work and risk systems for a real small business and run them for one month.
For a real business (yours or a client's), set up lean bookkeeping with a close checklist, invoicing with a reminder ladder, a thirteen-week cash sheet, a CRM with five stages and a weekly review, one task list, a file structure, three procedures, tested backups, a password manager with offboarding, and a compliance calendar. Run everything for one month and write a one-page review.
Design a three-step checkout with every state and all its copy, then run a usability test and revise.
Design a checkout flow: cart review, details, confirmation. Define each screen's job and primary action, build hierarchy in greyscale first, design all five states per screen, write all copy to the rules, then test with five people and revise the top three findings.
Build and deploy a SaaS with accounts, two plans with limits, subscription billing via webhooks, an admin customer view and the operational basics.
Ship a small SaaS: users and accounts, two plans with one enforced limit, checkout and webhook-driven activation with idempotency, a cancel-at-period-end flow, an admin customer screen with two audited actions, and the five operational services. Deploy it and take one real test payment.
Build a landing page with semantic structure, keyboard-accessible interaction, a fluid layout and measured performance.
Build a landing page for a product with a header, hero, feature grid, pricing cards, FAQ disclosures and a signup form. It must pass an accessibility audit, work by keyboard, adapt from 320px to 1600px with at most two media queries, and load with good paint metrics.
Take an untyped or loosely typed module and give it precise types: unions, boundaries, strict mode, no any.
Choose a real module with any, string-typed enums or unchecked JSON. Model its data with discriminated unions and literal types, add schema parsing at its boundaries, enable strict for it, and remove every any and cast. Keep behaviour identical, proven by tests or snapshots.
Build the browsing interface for a content library: list, search, filters, detail view, bookmarks — with data from a query library and tests for the key flows.
Build a React interface that lists resources, filters them by type and difficulty, searches by title, opens a detail view, and lets a user bookmark. Data comes from a mock API through a query library. Write tests for filtering, searching and bookmarking.
Build a small app where signed-in members browse resources, save bookmarks and see only what their plan covers.
Build an app with a public resource list (server-rendered), sign-in, an account area protected on the server, a server action to bookmark, and a plan check that hides full content from members without access. Deploy it with previews.
Take one manual process from map to live automation with error handling, a manual fallback and a measured saving.
Choose a recurring process. Map it from three real examples, build the automation one block at a time, add error routing and idempotency, test three samples and one deliberate failure, and run it live for two weeks with a log of runs, failures and time saved.
Configure instructions, three projects and a saved-prompt file for your real work, then measure the time saved over a week.
Write custom instructions, create three projects for your three most recurring areas of work with files and instructions, and keep a file of five saved prompts with variables. Use the setup for a week and log time saved per task and edits made.
Prepare a real repository so Claude Code can work in it safely: conventions, tests, checks and a release script, proven by one test-first feature.
Take a repository you own. Write CLAUDE.md, ensure the test command runs green, add a release-check script that prints an evidence table, then implement one small feature test-first with the tool and ship it with the table in the pull request.
Set up a Claude project that reviews contracts against your checklist and produces a referenced summary, tested on three real contracts.
Build a project whose instructions define a review checklist (payment, termination, liability, IP, renewal), an output format with section references and quotes, and a rule that ambiguous clauses are quoted and listed for a lawyer. Test on three contracts; verify every reference.
Build a read-only agent that answers questions with cited findings, within a fixed budget, and passes an evaluation set.
Build an agent with two tools — search and fetch — that answers a question with findings (each with URL and date), inferences (each with reasoning), and gaps. Enforce a budget of fifteen fetches and forty steps in code. Validate the report shape in code and reject uncited findings. Evaluate on twenty questions.
Build five reusable prompts for one role's recurring tasks, each with variables, a reference run and a ten-case test set.
Choose one role (support, sales, marketing, ops). Identify its five most repeated writing tasks. For each, produce a prompt template with variables, a filled reference run, and a ten-case test set with rubrics and a recorded pass rate. Deliver as a folder someone in that role could use on Monday.
Write a two-page guide for your team on what your chosen assistant does well, where it fails, and the rules for using it, with every claim tested by you.
Produce a two-page internal guide for a real team: what the assistant is, three tasks it does well with evidence from your own tests, three ways it fails with examples you produced, and five rules for use. The guide must not contain a claim you did not test yourself.
Milestone 1: Test the strengths
Milestone 2: Reproduce the failures
Any assistant
Instant delivery after paymentLicence stated on every productRe-download anytime from My LibrarySecure checkout — Polar or FungiesNew products every week
Instant delivery after paymentLicence stated on every productRe-download anytime from My LibrarySecure checkout — Polar or FungiesNew products every week
