Get a custom industry briefing every morning.

What the AI will do

• Investigates all the email and document sources it can access, along with relevant messages, meetings and upcoming work, to learn what you need to know.
• Researches the best information sources for those needs, shows you what it discovered, and produces a briefing that explains why each story matters to you.
• Builds in quick feedback on every story and retains your preferences, so each edition can improve in relevance, length, clarity and depth.

What you can provide

• Access to your email and work documents, where available; existing reading lists or briefings are welcome but not required.
• Any topic you already know you want covered or excluded.

Where you'll take part

• Correct what the AI infers about your interests and information needs.
• Choose how often, at what time and where to receive the briefing.
• React to individual stories: useful or not, longer or shorter, clearer or more in-depth.

What you'll get

A first personalized briefing and an editable briefing routine that learns from your story feedback, with your chosen delivery arrangement tested where the available tools support it.

Activity context

The best briefing knows that a small regulatory announcement could matter more to you than the day's biggest headline. Have AI look through your email and documents to work out what you need to know, then find the best sources for it. Tell it which stories helped, which missed, and whether you wanted more depth or fewer words. Each edition should become more like the briefing you would have made for yourself if you had the time. Morning is just one possibility: choose whatever frequency and delivery time suit you.

For the user

Start in an AI with access to your work records, or supply whatever rough material is available. The AI will map what it can search, investigate your needs, reveal what it found and make a first edition. You will correct its understanding, choose the delivery arrangement and react to a few stories. The AI should do the source research and the setup work its tools allow.

Guidance for the AI

These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result.

Start with the activity and the user's context

Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear.

Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them.

Check the capabilities needed for this work

Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit.

When useful access is missing:

• Explain what you cannot reach and how that material would help.
• Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot.
• If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response.
• Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available.

Ask focused questions and respect the answers

Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful.

Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements.

Keep evidence and independent contributions distinct

Distinguish evidence, inference, and missing information. Check your work against the source material.

Use independent agents when your system actually supports delegation and independent investigation or review would improve the work:

• Give each agent a bounded brief and the context appropriate to its role.
• Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations.
• Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately.

Produce, check, and revise the result

Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them.

Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result.

Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized.

Your AI prompt

Discover what I need to know

Create a personalized industry briefing that becomes more useful with every edition. Begin by discovering my information needs from my work, then research the best sources for those needs. The title mentions mornings; my actual frequency, time, time zone and delivery place are mine to choose.

Inspect and test the accessible email accounts, document repositories, chat sources, calendar and meeting transcripts. Establish the full available source inventory and any date or search restrictions. Investigate across all accessible email and document sources in manageable passes, rather than basing the briefing on one supplied memo or asking me to assemble a reading packet. Use recent correspondence and documents to discover projects, customers, competitors, markets, technologies, jurisdictions, recurring questions and people whose work I need to understand. Follow those terms into earlier strategy documents and related exchanges where they clarify a live issue. Inspect upcoming meetings and deadlines for needs with an immediate use.

Keep a coverage ledger with repositories searched, date ranges, completed batches and remaining gaps. Search access or a sample of hits is not the same as reading every document; report what you actually inspected. For a large corpus, delegate bounded source or business-area reviews to agents where available, with named inputs, evidence references, inferred needs, uncertainties and a completion criterion. Reconcile overlaps and contradictory needs. Otherwise use successive passes and preserve handoffs so the work fits the available session.

If a connection fails or a source is missing, investigate its current documented setup and export options, try what is available, and ask for the specific user action required. Proceed with the accessible record while clearly marking gaps. Supplied documents or pasted exchanges can support a limited first edition; do not describe that as an investigation of my whole working context.

Reveal the findings and choose the editorial brief

Show me what you learned in a concise account: my apparent responsibilities and active priorities, the questions I need answered, the topics I may be overlooking, and the work evidence behind those inferences. Distinguish things I explicitly requested from your suggestions. Ask the most useful unresolved question and wait where it affects topic selection. Let me correct the map without requiring a long questionnaire.

Create an editable editorial brief capturing confirmed needs, priorities, vocabulary, desired reading length, exclusions and important upcoming events. Keep uncertain interests labeled as provisional. This brief will evolve; do not freeze today's projects into permanent interests.

Find and assess the best information sources

Research current sources suited to each confirmed need. Look beyond a generic news search: original announcements and filings, regulators, standards bodies, research publications, industry data, specialist reporters and analysts, and relevant company or competitor material. Assess directness, expertise, timeliness, access and how much genuinely useful information each adds. Separate independent analysis from marketing and commercial claims. Use primary sources to verify material factual claims where possible.

Build a concise source map tying each source to the question it helps answer and noting relevant access limits. Read the actual items you use; a search snippet is a lead, not completed source inspection. Check publication dates and the date of the underlying event. Where a claim is disputed or has consequences for my decisions, trace it to the original evidence and show the uncertainty. Do not send confidential workplace material as public search text; translate internal needs into general research questions.

For substantial coverage, assign agents distinct topics with the confirmed editorial brief, source criteria, date window and required evidence. Require an account of what they inspected and could not access. Synthesize only after those bounded assignments finish, and resolve conflicting reports against the sources. Use separate passes if agents are unavailable.

Produce the first edition

Write a first briefing for the chosen scope. Lead with the developments most relevant to my work, including important small changes that broad headlines miss. For each story provide:

• A clear headline and the substance of what changed.
• Why it matters to a specific confirmed responsibility, project or question of mine.
• What the evidence establishes, and any material uncertainty.
• Direct links to the original sources and relevant dates.
• A stable, simple story identifier I can use when giving feedback.

Use my preferred length if known and a concise first edition otherwise. Avoid repeating the same underlying story through several articles. Separate new developments from useful background. Where nothing meaningful changed, say so instead of filling space. Reveal worthwhile findings outside the initial topic map as suggestions I can accept or dismiss.

Have a separate reviewer, where available, inspect the actual edition and its evidence for unsupported claims, weak relevance and unclear explanations without receiving your defense of the draft. Fix substantive findings. With no separate reviewer, perform a distinct source and readability pass and be honest that it shares context.

Make story feedback easy and persistent

Include a frictionless way to react to each story using the actual delivery tool: working reaction controls if supported, or simply “2 useful; 3 not useful; 4 shorter and clearer.” Support useful/not useful, longer/shorter, clearer/more in-depth and free-text comments. Explain that these reactions teach the briefing what I want; no response means no new preference evidence.

Ask for feedback on the first edition and apply what I supply. Retain the story identifier, topic, source, edition date, my exact reaction and the resulting editorial preference in an editable record that later runs can load. Distinguish a one-story reaction from a general rule; ask a brief question when the interpretation matters. Show a small revision so I can see that the feedback was understood. If I do not want to review now, keep the feedback route available and proceed using the confirmed preferences.

Use the record in every subsequent edition to revise story selection, source mix, explanation style, length and depth. Carry explicit preferences forward, reduce unwanted repetition and periodically refresh the work context so the briefing changes with my responsibilities. Let me view, correct or reset those preferences. Never claim to remember them across sessions unless a retained record is actually available.

Set the delivery and verify the whole loop

Ask how often I want it, at what time and in which time zone, and where I want to receive it. I may want weekdays, several times a day, weekly, a custom schedule or on demand. Inspect the actual scheduling, fresh-source retrieval, storage, feedback capture and delivery capabilities; research current official documentation as needed. Build the simplest arrangement the available tools support, carrying out the authorized setup and asking only for necessary decisions or access actions.

Every run must load the editorial brief and feedback, determine the new coverage window, refresh relevant work context, inspect sources, deduplicate against previous editions, write and verify the briefing, deliver it, and save the edition and updated state. Persist the retrieval checkpoint only after the edition is successfully recorded; make delivery failure visible rather than silently skipping a period. Keep the stored material proportionate to what the routine needs.

Test a complete first delivery: confirm the destination, readable edition and stable story identifiers. If I provide feedback, verify the captured reaction, saved preference and an example of its effect on the next edition. If I defer feedback, leave the route available and clearly mark capture and adaptation as awaiting my first reaction; do not require a reaction to finish the usable first edition or invent one to pass the test. Check the actual enabled schedule and show how to change or stop it. If an automated component cannot be provided, give me the finished briefing, retained editorial brief, source map and ready-to-run repeat instructions, with the missing component plainly identified. Do not call a proposed schedule active or a static file a working feedback system.

Finish with the briefing itself, a brief statement of the delivery arrangement and feedback method, and the actual tested state. Leave me able to use it immediately.

Done when

You have a personalized first edition and know how to comment on each story. If you supplied feedback, you can see it change the briefing; otherwise the feedback route is ready and its adaptation test is clearly pending. Your preferred cadence and delivery route are either operating and tested or clearly identified as a manual routine with complete reusable instructions.

These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.