[ { "id": "inbox-priority-review", "title": "Triage your inbox without automatically replying", "description": "Separate messages needing a decision from reference material and low-priority mail.", "category": "Customer work", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A source-linked inbox review with action items, deadlines, and unsent reply drafts.", "inputs": [ "A limited date range or mailbox label", "Your priority rules and examples of messages that matter" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Review [mailbox scope] using [priority rules]. Return messages requiring my decision, deadlines explicitly stated in the email, unanswered questions, and optional unsent reply drafts. Link to the original message and distinguish an explicit commitment from an inference. Do not send, archive, delete, unsubscribe, or change labels. Ask for clarification when the sender's intent is unclear.", "steps": [ "Connect only the mailbox access you intend ChatGPT to use, or provide approved exports.", "Define the time range and examples of high-priority messages.", "Compare the first brief to the original messages, including excluded items.", "Adjust rules and use an approval step for any later actions." ], "review": "Check deadlines, recipients, commitments, and any message the assistant may have missed.", "limits": [ "Mail access and available actions depend on your configured account.", "A large inbox demonstration does not prove reliable triage for every mailbox.", "Adapted from a Muse workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Missed important messages, false priorities, and corrections to draft replies.", "sources": [ { "title": "Muse email demonstration", "url": "https://x.com/Muse/status/2097440088812662803", "author": "Muse, featuring Casey and Hal", "note": "September 8, 2026. Official demonstration of working with a large email history. Our priority rubric and draft-only prompt are editorial additions. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "Vendor example" }, { "title": "Introducing Muse", "url": "https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/", "author": "Meta", "note": "Official introduction of the personal agent. Connector access, account availability, and action permissions still need to be configured.", "kind": "Official documentation" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "prospect-research-sheet", "title": "Research a small prospect list with evidence", "description": "Build account briefs you can verify before using them for outreach.", "category": "Sales", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A spreadsheet-ready account list with fit evidence, sources, and unsent outreach ideas.", "inputs": [ "Your customer profile and exclusion criteria", "A small approved account list or research scope" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Research [accounts or bounded search scope] against [customer profile]. Return company name, domain, relevant public facts, dated source URLs, and fit reasoning. Use unknown for missing details. Deduplicate records and separate factual evidence from an outreach hypothesis. Draft an optional short opener only for reviewed accounts. Do not guess private contact information or send messages.", "steps": [ "Specify company-level criteria and excluded categories.", "Start with a small research batch that can be manually checked.", "Review source URLs and company identity before exporting results.", "Have the owner approve any outreach and separately verify contact details." ], "review": "Check company matches, source dates, fit reasoning, and every personalized claim before use.", "limits": [ "The creator’s CSV report does not prove research completeness.", "Website access or third-party enrichment may require separate tools and subscriptions.", "Adapted from a Muse workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Verified fields, duplicate records, and reviewer-accepted account briefs.", "sources": [ { "title": "Prospect research with Muse", "url": "https://x.com/smallchris/status/2098373287453601935", "author": "Chris Small", "note": "September 11, 2026. Reports account research, CSV enrichment, and outreach drafts. The public excerpt does not establish completeness or accuracy. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "Practitioner report" }, { "title": "Introducing Muse", "url": "https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/", "author": "Meta", "note": "Official introduction of the personal agent. Connector access, account availability, and action permissions still need to be configured.", "kind": "Official documentation" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "supplier-comparison", "title": "Compare supplier quotes before you order", "description": "Put prices, lead times, minimum orders, and missing terms in one comparison.", "category": "Purchasing", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Intermediate", "evidence": "Editorial adaptation", "output": "A component sourcing table with exact part identifiers, supplier URLs, stock, quantity, cost and unresolved substitutions.", "inputs": [ "Bill of materials with exact part identifiers and required quantities", "Approved suppliers, delivery destination, budget, and deadline" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Research [bill of materials] against [approved suppliers and requirements]. For each line, list exact manufacturer and part number, supplier URL, quantity available, unit and extended cost, shipping and tax assumptions, lead time, and return terms. Separate exact matches from proposed substitutes and flag unclear specifications. Reconcile quoted quantities to the original list. Do not create supplier accounts, disclose confidential design details, place orders, or make payments. Return a sourcing table and unresolved questions for approval.", "steps": [ "Normalize part identifiers, specifications, and quantities in the bill of materials.", "Find exact supplier matches and record time-sensitive stock and prices.", "Compare landed cost assumptions, lead times, and any proposed substitutions.", "Have the project owner review compatibility and approve procurement separately." ], "review": "A technically qualified owner verifies component compatibility, authenticity, stock, quantities, and shipping constraints.", "limits": [ "The visible X excerpt reports research and procurement across 14 suppliers but is truncated. Do not claim transaction verification details or achieved time savings without the full source. This is a personal report, not a reproduced test.", "The prompt, setup steps, and review criteria are editorial adaptations; this directory has not reproduced the reported outcome.", "Adapted from a Muse workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Exact matches sourced, uncertain substitutes identified, and missing or duplicated line items.", "sources": [ { "title": "Compare suppliers for a hardware bill of materials — original X report", "url": "https://x.com/justinbojarski/status/2099630878993387556", "author": "Bojo.io", "note": "2026-09-14. Verified original author and public excerpt via X oEmbed on 2026-10-12. The visible X excerpt reports research and procurement across 14 suppliers but is truncated. Do not claim transaction verification details or achieved time savings without the full source. This is a personal report, not a reproduced test. Guide prompt and steps are editorial additions. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "Practitioner report" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "business-expense-audit", "title": "Find questions worth asking in business expenses", "description": "Group recurring spend and flag anomalies without calling every expense waste.", "category": "Money & reporting", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "An expense-review table with source rows, owner questions, and possible investigation areas.", "inputs": [ "Redacted expense export for a defined period", "Known vendors, cost centers, and contract context" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Review [expense export] for [period]. Preserve source row IDs and reconcile totals. Group vendors consistently, identify duplicate-looking charges, recurring commitments, unusual changes, and missing descriptions. Treat each as a question for the owner, not proof of waste or fraud. Do not change accounting records, cancel contracts, move money, or infer tax treatment. Return a discrepancy list and the reconciliation.", "steps": [ "Choose a reporting period and remove unnecessary payment details.", "Reconcile the imported rows to the accounting export total.", "Review grouped vendors and exceptions with the relevant budget owners.", "Document confirmed findings and handle accounting changes through the normal approval process." ], "review": "Finance checks duplicate assumptions, currencies, refunds, timing differences, and contractual commitments.", "limits": [ "Different transaction dates may represent legitimate installments.", "This is an investigation aid, not an audit opinion or tax classification.", "Adapted from a Grok Bot workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Reconciled totals and exceptions confirmed or dismissed by owners.", "sources": [ { "title": "Weekend bot wins supplied by the site owner", "url": null, "author": null, "note": "The pasted report mentions bill, shopping, subscription, expense, and travel tasks. Original author, date, and post URL are unverified; no savings are treated as established results. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "User-supplied report" }, { "title": "Grok Bot overview", "url": "https://docs.x.ai/grok-bot/overview", "author": "xAI", "note": "Official overview of the bot environment, browser, files, connectors, and approvals. It does not validate the reported savings in the supplied anecdote.", "kind": "Official documentation" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "commitment-tracker", "title": "Track commitments across Slack, Notion, and email", "description": "A reviewable commitments register with owner, due date, source link, and completion evidence.", "category": "Operations", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Intermediate", "evidence": "Editorial adaptation", "output": "A reviewable commitments register with owner, due date, source link, and completion evidence.", "inputs": [ "Selected Slack channels, Notion pages, and email date range", "Rules for what counts as a commitment and who owns each item" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Review [approved message and task scope] for commitments I explicitly made during [date range]. Return a table with task, owner, stated deadline, original quote or source link, status, and evidence of completion. Distinguish promises from ideas or requests. Deduplicate repeated discussions. Mark unclear owners and dates as unknown. Do not message colleagues, assign work, or write to Notion yet; show the proposed changes for review.", "steps": [ "Choose a narrow source scope and date range.", "Extract commitments with links to the original messages.", "Review uncertain wording, duplicates, owners, and deadlines.", "Approve any Notion updates separately and close items only when completion evidence exists." ], "review": "Check whether the user actually committed, whether the deadline was explicit, and whether a later message changes the obligation.", "limits": [ "One practitioner report; no measured accuracy or time saved. A Notion integration is reported by this user, not verified as universally available.", "The prompt, setup steps, and review criteria are editorial adaptations; this directory has not reproduced the reported outcome.", "Adapted from a Grok Bot workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Missed commitments, false task detections, duplicates, and unsupported completion marks.", "sources": [ { "title": "Track commitments across Slack, Notion, and email — original X report", "url": "https://x.com/samlambert/status/2099584402045210992", "author": "Sam Lambert", "note": "2026-09-14. Verified original author and public excerpt via X oEmbed on 2026-10-12. One practitioner report; no measured accuracy or time saved. A Notion integration is reported by this user, not verified as universally available. Guide prompt and steps are editorial additions. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "Practitioner report" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "subscription-review", "title": "Review recurring software and service costs", "description": "List business subscriptions, renewal dates, owners, and questions before deciding what to keep.", "category": "Money & reporting", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "An inventory of possible subscriptions with evidence, renewal details, and an owner decision column.", "inputs": [ "Redacted recent statements or approved billing exports", "Known subscriptions and services you want to keep" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation.", "prompt": "Help me with this small-business task using only the sources I provide or explicitly connect. Review [billing exports] for possible subscriptions. Group repeated merchants, preserve source rows, and distinguish confirmed recurring charges from guesses. List amount, frequency, latest charge, known renewal date, and what needs checking. Create keep, investigate, and possible-cancel columns without making the decision for me. Do not log in, send cancellation requests, or change services without separate approval.", "steps": [ "Use a small set of redacted statements and retain unchanged originals.", "Match recurring merchant patterns to actual service accounts.", "Check renewal, cancellation, refund, and business-dependency details.", "Decide service by service and verify any later cancellation confirmation." ], "review": "Confirm the service owner, renewal terms, shared users, and whether cancellation removes data or access.", "limits": [ "Merchant names do not always identify the service correctly.", "A recurring charge can be necessary or shared; low usage is not proof it should be canceled.", "Adapted from a Grok Bot workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account." ], "measure": "Subscriptions correctly identified, uncertain charges resolved, and confirmed decisions.", "sources": [ { "title": "Weekend bot wins supplied by the site owner", "url": null, "author": null, "note": "The pasted report mentions bill, shopping, subscription, expense, and travel tasks. Original author, date, and post URL are unverified; no savings are treated as established results. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source.", "kind": "User-supplied report" }, { "title": "Grok Bot overview", "url": "https://docs.x.ai/grok-bot/overview", "author": "xAI", "note": "Official overview of the bot environment, browser, files, connectors, and approvals. It does not validate the reported savings in the supplied anecdote.", "kind": "Official documentation" }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ], "reviewed": "2026-10-17" }, { "id": "x-lead-discovery", "title": "Find people on X asking for what your product does", "description": "Review a short list of recent buying signals every four hours, with direct post links, fit reasons, and helpful reply drafts.", "category": "Sales", "tool": "ChatGPT", "toolUrl": "https://chatgpt.com/", "level": "Intermediate", "evidence": "Editorial adaptation", "reviewed": "2026-10-17", "output": "A ranked shortlist of posts from the last 24 hours, with author handles, intent labels, relevance scores, and reply drafts. Each report ends with “What needs my attention?” and up to three strongest opportunities.", "inputs": [ "Product description, target customer profile, and clear exclusions", "Competitor names, customer problem phrases, and examples of relevant and irrelevant posts", "A persistent log of previously reported post IDs and handles, a report destination, and a timezone for the four-hour schedule" ], "connections": "Start with approved files or pasted excerpts in ChatGPT. If using connected apps, confirm access and permissions first. For recurring monitoring, use a configured task or dot; a single chat prompt does not create a running automation. X search availability is not assumed; you can paste public post excerpts and links instead.", "prompt": "Use the verified public post links I supply. Help me with this small-business task using only the sources I provide or explicitly connect. Act as a lead discovery agent for [PRODUCT].\nProduct description: [WHAT YOUR PRODUCT DOES]\nTarget customers: [WHO YOUR PRODUCT HELPS]\nCompetitors and exclusions: [COMPETITORS, GEOGRAPHY, AND EXCLUSIONS]\n\nFor each run of a separately configured four-hour schedule, search X for posts published within the last 24 hours. Record the run time and timezone. Focus on five types of posts:\n1. Requests for recommendations for products like mine.\n2. Complaints about problems my product solves.\n3. Frustration with competitors or requests for alternatives.\n4. Descriptions of workflows my product could improve.\n5. Explicit requests for help with something my product does.\n\nFor every relevant post, provide the original post text available to you and its direct link, the author's X handle, publication time, a specific fit reason tied to the product description, intent (actively looking, exploring alternatives, or expressing frustration), a relevance score from 1–10 with a one-line reason, and a suggested natural reply that helps the person. If no intent label fits, say unclear rather than forcing one. Reply drafts must make only supported product claims and must not imply that I am an unaffiliated customer.\n\nPrioritize explicit requests for a solution over general conversation. Use 8–10 for an explicit need and strong customer fit, 5–7 for a relevant problem with uncertain purchase intent, and 1–4 for weak fit; scores are judgment calls, not purchase probabilities. Exclude promotions, apparent bots, irrelevant mentions, and posts older than 24 hours. Bot detection is uncertain; flag ambiguous cases rather than claiming certainty. Never invent posts, links, user details, or quoted text. If the original cannot be verified, leave it out of the lead shortlist and note the access limitation.\n\nCheck [PERSISTENT REPORTED-LEAD LOG] before reporting. Do not repeat previously reported posts or leads. Group multiple posts from the same author into one opportunity. Record post IDs, handles, source times, and report times for leads included in a successfully delivered report. If the log cannot be read or written, disclose that and do not claim deduplication worked.\n\nKeep the report short, with a maximum of [10] new opportunities, highest intent first. A small or empty report is acceptable; do not pad it. End with “What needs my attention?” and highlight up to three most promising opportunities, or state that no qualifying new opportunities were found. Report incomplete searches or failed runs separately from zero matches.\n\nDo not reply, follow, DM, or otherwise contact anyone. I will review each opportunity and decide whether to respond.", "steps": [ "Define your product, customer profile, exclusions, and search phrases for all five signal types; choose a small report limit.", "Run one manual search and check each post, timestamp, fit reason, and reply draft against the original context.", "Set up a persistent reported-lead log and confirm that a repeated sample run does not resend the same posts or people.", "Configure the four-hour schedule and report destination in your own environment. Review the report before any response and inspect failures or empty runs." ], "review": "Open each original post before acting. Check that the request is current, the product actually fits, and the draft offers useful help without unsupported claims or disguised affiliation. A relevance score does not prove buying intent. All replies, follows, and DMs remain manual decisions.", "limits": [ "The excerpt’s author, original URL, and publication date are unverified. It describes an experiment and a suggested prompt, not measured lead or revenue results.", "X search coverage and account access can be incomplete. Freshness depends on checking post timestamps, not when a search engine indexed the page.", "A prompt alone does not create a scheduler or durable memory. Reliable deduplication requires a persistent log across runs.", "Intent scoring and bot detection are fallible. A complaint or workflow description is not necessarily a purchase request.", "Adapted from a Grok Bot workflow; this guide has not been reproduced with ChatGPT. Availability and integrations depend on your account.", "X search is not assumed to be available in every agent. Provide verified post links or configure a supported search tool first." ], "measure": "New opportunities accepted after review, duplicate and stale-post rates, unsupported fit claims, and conversations that progress after a human chooses to respond.", "sources": [ { "title": "X lead-discovery prompt — user-supplied excerpt", "url": null, "author": null, "kind": "User-supplied report", "note": "The supplied excerpt proposes four-hour X searches across five buying-signal types, a 24-hour freshness window, relevance scoring, deduplication, helpful reply drafts, and human review before any contact. The original author, URL, and date are not verified. This guide adapts the prompt and adds implementation checks; no conversion or revenue outcome has been established. This guide adapts the idea for a small-business owner; any reported outcome belongs to the original tool and source." }, { "title": "OpenAI use-case library", "url": "https://learn.chatgpt.com/use-cases", "author": "OpenAI", "kind": "Official documentation", "note": "Official examples span ChatGPT, Work and Codex. This small-business prompt is our adaptation, not a demonstrated customer result." } ] }, { "id": "customer-call-followup", "title": "Prepare the follow-up after a customer call", "description": "Turn the actual conversation into a reply draft, open questions and proposed CRM notes.", "category": "Sales", "tool": "ChatGPT Work", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A customer email draft and internal next-step list linked to the meeting record.", "inputs": [ "An approved meeting transcript or contemporaneous notes", "Customer identity, existing proposal and relevant prior commitments" ], "connections": "Uploaded notes can support the first pass. Reading live Zoom, email or CRM context requires each supported connection and its permissions.", "prompt": "Use [meeting record] and [approved account context] to prepare a follow-up for [customer]. Separate agreed scope from proposals, identify open questions, and list explicit owners and dates without inventing either. Draft a short customer email and separate internal CRM notes. Preserve references to the supporting transcript or notes. Keep internal concerns out of the customer draft. Show both for approval before sending messages, creating tasks or updating any record.", "steps": [ "Choose the exact customer meeting and relevant account context.", "Extract agreements, unresolved questions and stated next steps.", "Draft a customer reply separately from internal notes.", "Check the record and approve each intended message or system update." ], "review": "Verify names, commitments, dates and any price or scope mentioned in the reply.", "limits": [ "Transcripts can omit context or misidentify speakers.", "A suggested next step is not a commitment accepted by the customer." ], "measure": "Follow-ups reviewed, commitments correctly captured and corrections needed before sending.", "sources": [ { "title": "Turn meetings into follow-ups", "url": "https://learn.chatgpt.com/use-cases/zoom-meeting-follow-ups", "author": "OpenAI", "note": "Official ChatGPT workflow covers transcript-based email drafts, account plans and proposed CRM updates. This service-business version and its prompt are original adaptations.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "weekly-owner-report", "title": "Explain this week's sales and operating changes", "description": "Give the owner a short report with reconciled numbers, changes worth reviewing and unanswered questions.", "category": "Money & reporting", "tool": "ChatGPT Work", "toolUrl": "https://chatgpt.com/", "level": "Intermediate", "evidence": "Editorial adaptation", "output": "A one-page operating report with a source table and review questions.", "inputs": [ "Sales, cost and operating exports for comparable periods", "Metric definitions, targets and owner notes explaining known events" ], "connections": "Start with approved files. Live refreshes require separately connected data sources and a confirmed schedule.", "prompt": "Prepare a weekly owner report from [approved exports] using [metric definitions]. Reconcile source totals before comparing periods. Show the few material changes in sales, costs and operations, with a source reference for every number. Separate explanations supported by owner notes from hypotheses to investigate. List missing data and inconsistent definitions. Return a draft and supporting calculations; ask for approval before sharing the report, editing source records or creating a recurring update.", "steps": [ "Choose comparable reporting periods and agree on metric definitions.", "Reconcile imported totals and identify missing or inconsistent fields.", "Draft the material changes with source references and clearly labeled hypotheses.", "Have the owner verify the calculations and approve distribution or a recurring schedule." ], "review": "Check period cutoffs, refunds, accrual versus cash figures and whether an apparent change comes from a data-definition change.", "limits": [ "A weekly change may be seasonal or caused by incomplete data.", "An agent-generated explanation is not proof of causation or a reliable forecast." ], "measure": "Figures reconciled, unsupported explanations removed and decisions informed by verified changes.", "sources": [ { "title": "Prepare a business review", "url": "https://learn.chatgpt.com/use-cases/monthly-business-review-narrative", "author": "OpenAI", "note": "Describes preparing source-backed business reviews from workbooks, KPI definitions and owner notes. This smaller weekly owner report is an editorial adaptation, not a tested performance claim.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "morning-owner-brief", "title": "Start the day with the customer work that needs attention", "description": "Bring deadlines, unanswered customer messages and today's appointments into one short briefing.", "category": "Operations", "tool": "ChatGPT Work", "toolUrl": "https://chatgpt.com/", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A dated priority list with source links, missing context and optional reply drafts.", "inputs": [ "Approved inbox, calendar and job-tracker sources", "Business hours, priority rules and escalation preferences" ], "connections": "Connect the specific supported email, calendar and tracker tools needed. A schedule is active only after it has been saved and confirmed.", "prompt": "Review [approved sources] for today's customer work in [time zone]. Return no more than five items needing attention: appointments to prepare for, customer replies, deadlines at risk and decisions waiting on me. Check latest replies so resolved work is not repeated. Link each item to evidence and name unavailable sources. Draft replies where helpful, but ask for approval before sending, changing appointments or updating jobs. Propose a recurring schedule only after this first brief is reviewed.", "steps": [ "Choose the exact sources and define which issues deserve attention.", "Prepare one briefing using a bounded date range.", "Review missing sources, duplicate items and priority judgments.", "Approve a schedule and notification rules after the trial brief is useful." ], "review": "Confirm the brief reflects the latest replies and does not treat tentative appointments as confirmed work.", "limits": [ "Disconnected tools and private threads can leave gaps in the briefing.", "A reminder does not complete the underlying task or guarantee a deadline will be met." ], "measure": "Relevant items surfaced, resolved items wrongly repeated and missed commitments found in review.", "sources": [ { "title": "Set up a work chief of staff", "url": "https://learn.chatgpt.com/use-cases/daily-work-brief", "author": "OpenAI", "note": "Official workflow describes recurring cross-tool briefings and draft replies. Our customer-work focus and five-item limit are editorial choices; no time savings are claimed.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "event-venue-headcount", "title": "Compare event venues and keep the headcount current", "description": "Keep venue tradeoffs, confirmations and pricing questions together as an event takes shape.", "category": "Dots: ongoing work", "tool": "OpenAI dots", "toolUrl": "https://learn.chatgpt.com/docs/dots", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A venue comparison and dated headcount tracker with open questions.", "inputs": [ "Event dates, location, budget, capacity and accessibility needs", "Approved venue proposals and attendee confirmation sources" ], "connections": "Public venue research can use the dot's cloud browser. Private proposals and confirmations need the relevant connected app or a separate authenticated session.", "prompt": "Help plan [event] under [requirements]. Compare a bounded venue shortlist using current source links, total-cost assumptions, capacity and unanswered questions. Track confirmed, declined and pending attendees from [approved sources], retaining an as-of time. Show how headcount affects quoted costs without inventing a new quote. Ask for approval before emailing venues or attendees, requesting revised pricing, reserving space, sharing the attendee list or paying a deposit.", "steps": [ "Define the event requirements and approved research sources.", "Compare venues using current proposals and clearly marked unknowns.", "Track attendee responses and flag changes affecting the plan.", "Review availability and pricing, then approve any contact or reservation separately." ], "review": "Check the proposed date, capacity, accessibility, headcount and full terms before making a commitment.", "limits": [ "Listed venue prices may omit fees or be unavailable for the requested date.", "Connecting an app does not automatically establish a monitoring task or permission to contact people." ], "measure": "Confirmed venue facts, headcount discrepancies resolved and pricing questions closed.", "sources": [ { "title": "Meet dots", "url": "https://learn.chatgpt.com/docs/dots", "author": "OpenAI", "note": "Official documentation illustrates venue comparison and describes tracking offsite confirmations while asking before requesting revised pricing. The business-event checklist and prompt here are original adaptations.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "planning-deadline-watch", "title": "Catch changed deadlines in a planning channel", "description": "Review a defined channel and show which job dates or commitments need an owner's decision.", "category": "Dots: ongoing work", "tool": "OpenAI dots", "toolUrl": "https://learn.chatgpt.com/docs/dots", "level": "Intermediate", "evidence": "Editorial adaptation", "output": "A proposed checklist update and a short exception report with message links.", "inputs": [ "One approved planning channel and the current project checklist", "Check frequency, time zone, end date and notification thresholds" ], "connections": "A supported connected messaging service and access to the named checklist. Ask the dot to confirm whether scheduling or the requested event monitoring is supported.", "prompt": "Review [planning channel] against [project checklist] for changes to agreed deadlines and owners. Cite the exact message behind each proposed change and distinguish a suggestion from a confirmed decision. Prepare a read-only trial report first. Ask for approval before editing the checklist, messaging anyone or activating [schedule, time zone and end date]. Once approved, confirm what was saved and report only meaningful deadline risks, failed checks or decisions that need me.", "steps": [ "Name the exact channel, checklist and changes to watch for.", "Run a trial comparison and review the proposed updates.", "Approve a supported schedule and any limited checklist writes explicitly.", "Confirm the saved settings and inspect the first scheduled result for missed or repeated changes." ], "review": "Verify that a changed deadline was agreed by the relevant owner and that the schedule uses the intended time zone.", "limits": [ "Connecting a channel alone does not start monitoring.", "Event support depends on the connected service; unsupported triggers must not be presented as active." ], "measure": "Confirmed deadline changes detected, false alerts and checklist updates that need correction.", "sources": [ { "title": "Tasks and memory", "url": "https://learn.chatgpt.com/docs/dots/tasks-and-memory", "author": "OpenAI", "note": "Documents saved schedules, supported event monitoring and a planning-channel checklist example. Our trial-report and approval sequence is an editorial adaptation.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "customer-website-bug-review", "title": "Turn a website problem report into a reviewed fix", "description": "Reproduce a customer-facing issue, prepare a focused change and bring back evidence before release.", "category": "Dots: ongoing work", "tool": "OpenAI dots + Codex", "toolUrl": "https://learn.chatgpt.com/docs/dots", "level": "Advanced", "evidence": "Editorial adaptation", "output": "A reproducible issue report, proposed code change and relevant test or visual evidence.", "inputs": [ "A customer bug report with the affected flow and environment", "Authorized repository access, project setup and test requirements" ], "connections": "A configured Codex cloud environment or an authorized connected computer with the correct repository. Local work requires that computer online with the app open.", "prompt": "Investigate [customer website issue] in [authorized repository and test environment]. Reproduce the failure, record the relevant steps and distinguish observed behavior from guesses. Prepare the smallest suitable patch and run checks that exercise the affected flow. Return the diff, results and any remaining uncertainty. Ask for approval before publishing a branch or pull request, merging, deploying, modifying production data or contacting the customer.", "steps": [ "Confirm the affected flow and reproduce it in the approved environment.", "Trace the cause and prepare a focused local change.", "Run relevant tests and capture evidence of the corrected behavior.", "Review the change and approve publishing or deployment through the normal release process." ], "review": "A responsible reviewer checks scope, regressions and test evidence before a change reaches customers.", "limits": [ "A completed agent run is not proof that the issue was fixed or a deployment succeeded.", "Missing reproduction details or environment differences can prevent a reliable diagnosis." ], "measure": "Issue reproduced, affected checks passed and reviewer-confirmed resolution after release.", "sources": [ { "title": "Tasks and memory", "url": "https://learn.chatgpt.com/docs/dots/tasks-and-memory", "author": "OpenAI", "note": "Official documentation includes a mobile-checkout investigation with a draft PR and test evidence, and explains cloud versus local task requirements. This website workflow is an original adaptation, not an observed customer result.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" }, { "id": "customer-meeting-pack", "title": "Prepare for a customer meeting from email and documents", "description": "Find the agreed scope, recent changes and questions worth resolving before the next conversation.", "category": "Dots: ongoing work", "tool": "OpenAI dots", "toolUrl": "https://learn.chatgpt.com/docs/dots", "level": "Beginner", "evidence": "Editorial adaptation", "output": "A short internal meeting pack with source links, a suggested agenda and unresolved questions.", "inputs": [ "Customer identity, meeting purpose and date", "Approved email threads, proposal documents and previous meeting notes" ], "connections": "Separately connect supported email and document plugins. The dot's cloud browser does not inherit the owner's personal-browser sign-ins.", "prompt": "Prepare an internal meeting pack for [customer and meeting] using only [approved threads and documents]. Summarize agreed scope, recent developments, open questions and commitments with direct source references. Mark stale or conflicting information and separate internal concerns from material suitable for the customer. Suggest an agenda and possible follow-up wording. Ask for approval before sharing the pack, sending messages, changing the calendar or updating customer records.", "steps": [ "Name the meeting and the exact email and document sources.", "Check the latest available context and identify missing material.", "Prepare the internal brief, agenda and source-linked questions.", "Review sensitive details and commitments before approving any sharing or follow-up." ], "review": "Check that the brief uses the right customer, current proposal version and latest conversation, and keeps private internal details separate.", "limits": [ "Plugin access depends on the connected account, permissions and execution environment.", "The supporting meeting use case is documented for ChatGPT Work; this Dots version is a proposed adaptation, not a demonstrated Dots customer result." ], "measure": "Preparation questions resolved, source conflicts identified and incorrect commitments caught before the meeting.", "sources": [ { "title": "Connect computers and apps to your dot", "url": "https://learn.chatgpt.com/docs/dots/computers-and-apps", "author": "OpenAI", "note": "Establishes separately connected app access, cloud-browser sessions and supported email/document work. It does not establish this exact customer-meeting outcome.", "kind": "Official documentation" }, { "title": "Turn meetings into follow-ups", "url": "https://learn.chatgpt.com/use-cases/zoom-meeting-follow-ups", "author": "OpenAI", "note": "Official ChatGPT workflow supports source-based meeting drafts and review before downstream actions. Used as inspiration for this original Dots meeting-preparation guide.", "kind": "Official documentation" } ], "reviewed": "2026-10-17" } ]