Prepare for a 1:1 with a report or manager
Turn scattered notes about a direct report or manager into a focused 1:1 agenda with real questions, not a generic status-check list.
Last updated
Fill in the blanks
Assembled in your browser — nothing you type is stored or sent anywhere.
Your prompt
Model-agnostic — works in Claude, ChatGPT, Gemini or any AI chat.
Why this prompt works
- →Asking for an open-ended question per topic instead of a status check forces the model to convert observations into something a person can actually respond to, rather than a checklist that produces yes/no answers.
- →Splitting recognition into its own section stops 1:1 prep from defaulting to a list of problems, which is the single most common way these prompts drift into a performance-review tone.
- →The instruction to flag under-supported topics instead of inventing a question prevents the model from manufacturing a plausible-sounding concern from a one-line observation that doesn't actually support one.
Part by part
- a specific, open-ended question rather than a yes/no check-in
- The clause that changes the output most. "Is everything going well?" gets "yes" from the people most likely to be struggling.
- things worth recognizing or thanking them for
- Recognition is the section that gets skipped when an agenda is written under time pressure, which is why it has to be a named heading rather than an intention.
- one thing to ask about their side that hasn't come up yet
- Reserves space for the report's agenda rather than the manager's. Without it a 1:1 becomes a status meeting with a warmer name.
- Keep the whole agenda to what fits in 30 minutes
- A length constraint expressed as time, not item count. It forces prioritisation rather than producing twelve topics you will not reach.
- say so instead of inventing a generic question
- The guard. Generic questions read as filler to the person answering them, and they are worse than a shorter agenda.
Common mistakes
Help me prepare for a 1:1.With no observations, the output is a list of standard questions about workload and blockers. Those are answerable without knowing the person, which is exactly why they produce nothing useful.
What should I ask my report in our 1:1?Returns a generic question bank. The value of the exercise is turning your own scattered notes into specific questions, and a question bank skips the part that required the notes.
Summarise these notes into talking points.Produces a briefing for you to deliver rather than an agenda to discuss. A 1:1 where the manager talks through prepared points is a status update, and the structure here exists to prevent that.
Questions
What counts as a useful observation to paste?
Specifics with a time and a context — a piece of work that went well, a meeting where something seemed off, a decision they pushed back on. Impressions without an anchoring event produce questions that cannot be asked directly.
Should I share the agenda beforehand?
Usually yes. It gives the other person time to think, and it signals that the meeting is not an assessment. The one exception is a topic where a considered answer is less useful than an unprepared one.
Why is the recognition section separate?
Because it is the first thing dropped when an agenda gets long, and its absence is noticed cumulatively rather than in any single meeting. Making it a heading means dropping it becomes a visible choice.
Is it appropriate to use a model for this at all?
For structuring your own observations, yes. For generating the observations, no — a 1:1 agenda assembled from nothing you noticed is worse than turning up with none, because it performs attention rather than paying it.