Plan a lighting schedule before switching it on
Supply days, timezone and light names to draft a schedule with brightness and disable instructions for manual setup.
Supply days, timezone and light names to draft a schedule with brightness and disable instructions for manual setup.
Provide a city, available dates, preferences and budget. Read organizer or venue pages and list candidates with links and check times.
Supply a de-identified transaction file and date range. List merchants, dates, amounts, currencies, source rows and reasons for flagging repeats.
Supply your use, budget and must-haves to draft reusable buying criteria and a sourced comparison for your own decision.
Use a safely deployed OpenClaw setup with an authorized to-do interface and test list. Submit the item, then read back the record.
Text exercise: Provide redacted Home Assistant settings, logs, expected behaviour and the observed problem. Separate evidence, possible causes and open questions.
Text exercise: Supply a de-identified call transcript and CRM field definitions. Draft field values and follow-up copy using confirmed facts only.
Supply a de-identified report, currency and reporting window to compare each metric's inputs and formulas before interpreting results.
Provide a de-identified software inventory and public advisories. Preserve CVE identifiers, source links and original ratings in a review list.
Text exercise: Provide authorized discussion text, a research question and a date range. Draft themes, disagreements, counterexamples and sources.
Use a private test bot, rules, fictional characters and content boundaries. In a configured n8n workflow, record before-and-after state and test recovery.
Draft a sample chapter from your material, then edit it in Google Docs and export a PDF for layout review before expanding.
Text exercise: Provide fictional contact-form emails and tone requirements. Identify the proposed recipient with evidence and draft an anonymized topic summary.
Text exercise: Provide consented team updates, a time period and an audience. Draft a source-linked summary, marking missing updates without judging performance.
Using Claude where available: Supply an original article, audience and length limits to get draft posts with sources and open checks.
Give your region, genre, language and running-time preferences. Use accessible legal-platform pages to list titles, years, directors and sources.
Give the area, total budget, move-in date, requirements and permitted sources. Run one listing search and record rent, visible fees, original pages and unknowns.
Provide a scoped email list or authorize read-only mailbox access. Show representative messages, frequency and reasons for each suggestion.
Using Codex CLI where available: Supply authorized code and approved rules to generate test inputs and record reproducible failures.
Use Codex with a Rocq test project and proof obligation to prepare a minimal proof repair and checker logs through the test workflow below, then review it. Matching tools and permissions are required.
Use Codex with tempo, instruments and section requirements to prepare editable track and MIDI drafts through the test workflow below, then review it. Matching tools and permissions are required.
Use Codex with fictional inventory and authorized recipes to prepare a testable recipe-filter prototype through the test workflow below, then review it. Matching tools and permissions are required.
Use Replit Agent with fictional transactions and currency rules to prepare a ledger prototype and known-answer tests through the test workflow below, then review it. Matching tools and permissions are required.
Give a text AI tender requirements and a response draft to draft a requirements-and-gaps table with page references for human review. This is a drafting exercise; the source used Replit Agent.