Working with an assistant
Using agents · topic using-agents/assistant-work
The everyday uses of a chat assistant for knowledge work: finding and summarizing information, thinking a problem through out loud, producing and reshaping drafts, getting critical feedback on your own work, and working from files and images you supply rather than from the model's memory.
Concepts
- Finding information
- Using an assistant to locate, summarize and compare information, in documents you supply or, with a search tool, on the web. It is fast at orientation and at turning a long text into the three things you need to know. It is unreliable on specifics from memory alone, so ask for sources, supply the material yourself where you can, and check anything you will repeat. glossary
- Thought partner
- Using the assistant to think, not to answer: explain your problem, ask it to question your assumptions, list what you might be missing, argue the opposite view or lay out options. The value is in your own clearer thinking. Ask for challenge explicitly, because a model left to itself tends to agree with the framing you give it. glossary
- Drafting
- Getting a first version of an email, report, plan or message from a brief, then shaping it. Give the audience, the purpose, the points that must appear and the tone; ask for alternatives when you are unsure. The draft is raw material, not the product. The facts in it are yours to supply and check, and the final voice should be yours. glossary
- Critique
- Asking an assistant to review your own work: find weak arguments, unclear passages, missing steps or errors. Because models tend to praise, ask for problems only, ask it to grade against stated criteria, or tell it the text is someone else's. Critique is one of the safest uses, since you remain the author and judge every suggestion. glossary
- Working with files and images
- Giving the assistant a document, spreadsheet, screenshot or photo to work from instead of describing it. This grounds the answer in your material and unlocks tasks such as extracting a table from a picture or comparing two versions of a contract. Mind what is in the file before uploading; attachments are data disclosure just like pasted text. glossary
Links
- Builds on: Prompting
- Leads to: nothing yet
- Competencies drawing on it: Delegates a task to an agent and checks the result
Lessons
- Drafting a text and asking for critique (tutorial)
- Finding information with an assistant (tutorial)
- Thinking a problem through with an assistant (tutorial)
- Working from your own files and images (tutorial)
Your reference
Each lesson above adds its takeaways and its example here once you finish it. Your reference lists every lesson you have finished.
Drafting a text and asking for critique
Unlocks when you finish Drafting a text and asking for critique.
Takeaways
- A one-line request gets a draft for nobody in particular, and the model fills the gaps with invented facts.
- A drafting brief names the audience and purpose, the points that must appear, the tone and the length, and it ends with done-criteria written before the draft exists.
- The draft is raw material. Every fact in it is yours to supply or check, and the final voice is yours.
- Asked what it thinks of your text, an assistant tends to praise it. Feedback is more positive when the user says they wrote the text.
- Ask for problems only, against stated criteria, and call the text a colleague's. That framing leans the critique harsh, so you stay the judge of every finding.
Example
Write an email to my team about the new expense process.
Subject: Important update to our expense process
Dear team,
I am writing to let you know about some upcoming changes to the way wesubmit expenses. Starting next month, we will move to a new system thatmakes submitting claims faster and easier for everyone.
The new process will simplify approvals and reduce paperwork. Pleasereview the attached guidelines before you submit your next claim, andcontact [Name] in Finance if you have any questions.
Thank you for your cooperation.
Best regards,[Your name]Finding information with an assistant
Unlocks when you finish Finding information with an assistant.
Takeaways
- An assistant is fast at orientation in a text you supply, and unreliable on specifics from memory alone.
- "Summarize this" gives you a paragraph you can only check by reading the source again.
- A brief for finding information gives the assistant the material, marks it as the only source, and asks for the sentence behind each claim and for a plain "the text doesn't say" where there is none.
- Check the quotes. A quote that is not in the text, even when the fact is right, shows where the model joined or added something.
- A confident detail from memory is the most likely text for the question, and the model didn't look it up. Confidence in the prose says nothing about the fact.
Example
Summarize this.
[the guideline, pasted in full]
This guideline explains how employees of Brindlewick Instruments claim back expenses. Claims go through the intranet form within 14 days and need receipts, and claims older than 90 days are not paid. Payment comes with the next salary run. Only team leads have company cards. Meals on business trips are reimbursed against receipts up to a daily limit, and alcohol is excluded. Train is the default for shorter trips, flights require manager approval and must be economy, and taxis are only allowed when there is no public transport. Mileage in your own car is reimbursed at a fixed rate when no train is available. New hires can claim a one-off amount for a home office desk or chair within their first three months. Questions go to the finance team during their office hours on Tuesdays and Thursdays.
Thinking a problem through with an assistant
Unlocks when you finish Thinking a problem through with an assistant.
Takeaways
- A thought partner helps you think a choice through, and the value of the session is in your own clearer thinking.
- A request for a verdict gets your own framing back. The description decides which reasons the model can give, and assistants often match the view you have shown.
- A brief for thinking out loud states the problem and the facts without adjectives, names your lean as a lean, and asks for the case against it and for the questions you haven't answered.
- Ask for no verdict. The challenge comes only when the brief asks for it.
- Close the session by writing down the one point that changed your view, or the one question you still have to answer.
Example
I have two job offers and need to answer by Friday. Which one should I take?
[the two offers, as written above]
Based on what you describe, Offer A sounds like the better fit. You got along well with the team, the product excites you, and a 10 percent raise with mostly remote work is a real improvement. Offer B is stable, but a long commute three days a week and a slow, formal culture can tire you over time. Go with Offer A, and ask the startup a few questions about its plans before you sign.
Working from your own files and images
Unlocks when you finish Working from your own files and images.
Takeaways
- A description of a table is a lossy copy of it, and the loss is in the exact values. Attach the picture or the file, and let the model do the typing.
- The brief for an upload has the same four parts as any other. Name the attachment as the only source, say which values must be copied exactly, and give the model a way to say it can't read a cell.
- Check the result against the file, cell by cell. A model reads small, blurry or rotated text badly, and a transcription error is silent until you look.
- When comparing two files, say which kinds of difference matter and where the rest should go, and ask for the quote from both versions behind each one.
- An attachment discloses its whole content, the same as pasted text. Crop or copy out the part the task needs before you attach.
Example
I have a stock count table with six rows. The columns are item code, description, shelf, counted, expected and difference. The items are a torque wrench, a calibration weight, two digital calipers, a thermocouple probe and a spirit level. The item codes start with BI and a dash and a four-digit number. Can you give me the rows to paste into a spreadsheet?
Here is the table as tab-separated rows. I have filled in the codes and numbers as placeholders where you did not give them, so replace those with the real values.
Item code Description Shelf Counted Expected DifferenceBI-0001 Torque wrench B1 10 10 0BI-0002 Calibration weight B1 5 5 0BI-0003 Digital caliper B2 12 12 0BI-0004 Digital caliper B2 12 12 0BI-0005 Thermocouple probe B3 8 8 0BI-0006 Spirit level B3 4 4 0Sources
DLAI-3AI Prompting for Everyone, DeepLearning.AI (course)Academy claude-101Claude 101, Claude Academy (course)Academy introduction-to-claude-coworkIntroduction to Claude Cowork, Claude Academy (course)