120 Claude Prompts That Write Your Meeting Notes

<context> I need a clean action-item list from [meeting type, e.g. weekly product sync] so nothing slips through. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Attendees: [names and roles] - Today's date: [date] </inputs> <task> Extract every commitment as a table with columns: Owner, Action, Due Date, Priority, Confidence. Infer a due date only when clearly implied by the discussion and mark it [inferred]. Set Confidence to High/Medium/Low based on how explicit the commitment was. </task> <constraints> - One row per discrete action, no bundling - If no owner was named, put [unassigned] and flag it - Quote the trigger phrase from the notes in a final column for traceability </constraints> <format> Markdown table sorted by due date, then a short bullet list of anything ambiguous I should confirm. </format>

<context> After [meeting name] I want to send each owner a short note listing only the items they committed to. </context> <inputs> - Notes/transcript: [paste transcript or notes] - My name and role: [name, role] - Tone: [warm but direct] </inputs> <task> Group every action item by owner, then draft one email per owner containing only their items. Each email should open with one friendly line, list their actions with due dates, and close with a single clear ask to confirm. </task> <constraints> - Keep each email under 120 words - Never include another person's tasks in someone's email - Use the recipient's first name in the greeting </constraints> <format> For each owner: a heading with their name, a Subject line, then the email body. </format>

<context> I want to bulk-import this meeting's action items into [task manager, e.g. Asana]. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Target tool: [Asana / Jira / Linear / Notion] - Default project or board: [project name] </inputs> <task> Turn each action item into an importable task row. Map fields appropriately for the chosen tool: Task Name, Assignee, Due Date, Project, Priority, and a one-line Description. </task> <constraints> - Use ISO date format YYYY-MM-DD - Leave Assignee blank and flag it if no owner was named - Do not invent statuses the tool does not support </constraints> <format> A CSV code block with a header row, ready to paste or upload. Below it, note any field I must map manually in the tool. </format>

<context> This [project/initiative] meeting left ownership fuzzy and I need a RACI to clarify accountability. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Key workstreams or deliverables discussed: [list, or "infer from notes"] - Stakeholders: [names and roles] </inputs> <task> Build a RACI matrix where rows are deliverables or decisions and columns are stakeholders. Assign R, A, C, or I to each cell based on what was actually said. Ensure exactly one Accountable per row. </task> <constraints> - Where the notes do not specify a role, mark the cell [TBD] rather than guessing - Flag any deliverable with no clear Accountable owner </constraints> <format> Markdown RACI table, followed by a short list of gaps to resolve before work starts. </format>

<context> Several action items from [meeting] depend on each other and I need to see the sequence. </context> <inputs> - Action items: [paste action list or transcript] - Owners: [names] - Hard deadline for the overall goal: [date, if any] </inputs> <task> Identify which action items block or depend on others. Produce a dependency map showing predecessors and successors, then flag the critical path. Call out any item that is a bottleneck because multiple others wait on it. </task> <constraints> - Base dependencies only on logical or stated relationships, not assumptions - Mark inferred dependencies as [inferred] </constraints> <format> First a table: Item, Depends On, Blocks, Critical Path (Y/N). Then a simple text-based sequence diagram showing the order of work. </format>

<context> I want to know which action items from the previous meeting are still open going into this one. </context> <inputs> - Last meeting's action items: [paste prior action list] - This meeting's notes/transcript: [paste current notes] - Today's date: [date] </inputs> <task> Compare the two sets. For each prior action item, classify it as Done, In Progress, Still Open, or Dropped, citing the evidence from this meeting's notes. For Still Open items, note whether a new due date was given. </task> <constraints> - If an item was not mentioned at all, mark it Status Unknown rather than assuming Done - Preserve the original owner unless reassigned in the new notes </constraints> <format> Markdown table: Item, Owner, Prior Due Date, Current Status, Evidence, New Due Date. End with a short list of overdue items needing immediate attention. </format>

<context> I only want the action items that I personally own, ignoring everyone else's. </context> <inputs> - Notes/transcript: [paste transcript or notes] - My name and any nicknames: [name, nickname] - Today's date: [date] </inputs> <task> Find every commitment assigned to me, including ones phrased as "you should", "can you", or implied by context. Turn them into a personal to-do list with due dates and a suggested order to tackle them. </task> <constraints> - Include items where I implicitly accepted a task, even if my name was not said - Exclude anything owned by someone else - Mark any item where my ownership is ambiguous as [confirm I own this] </constraints> <format> A checklist using - [ ] items, ordered by due date, with priority labels. </format>

<context> This cross-functional meeting produced actions across [teams, e.g. eng, design, marketing] and I want them grouped by team. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Team mapping: [owner to team, or "infer from roles"] </inputs> <task> Cluster every action item under the owning team or function. Within each team, list items with owner, due date, and priority. Note any action that spans two teams and needs coordination. </task> <constraints> - If an owner's team is unknown, place the item under [Unassigned Team] and flag it - Do not duplicate an item across teams; place cross-team items once and label them shared </constraints> <format> A section per team with a sub-table of its actions, then a short "Cross-team handoffs" list. </format>

<context> I want my action items from [meeting] blocked onto my calendar so I actually do them. </context> <inputs> - My action items: [paste my actions or transcript] - My working hours and timezone: [e.g. 9-5 ET] - Today's date and any fixed deadlines: [details] </inputs> <task> For each action, estimate the focused time it needs and propose a calendar hold before its due date. Avoid stacking everything on one day; spread holds sensibly across available time. </task> <constraints> - Respect stated working hours and do not schedule on weekends unless I say otherwise - Use realistic durations; flag anything that needs more than a 90-minute block to be split </constraints> <format> A table: Action, Estimated Duration, Suggested Day, Suggested Time Block, Due Date. Then the same holds as ICS-style VEVENT lines I can adapt. </format>

<context> Some action items are stuck on things outside the team's control and I need an escalation list for [leader/sponsor]. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Who can unblock things: [names and what they own] </inputs> <task> Identify every blocker raised: missing decisions, resource gaps, external dependencies, or approvals. For each, state what is blocked, why, who needs to act, and the impact if it is not resolved. Recommend the escalation owner. </task> <constraints> - Only list genuine blockers, not normal in-progress work - Rank by impact and urgency - Be specific about the exact decision or resource needed </constraints> <format> A prioritized table: Blocker, Blocked Item, Owner Needed, Impact if Unresolved, Escalate To. Then a 3-sentence summary I can paste into a message to leadership. </format>

<context> After [meeting] I want a single clear table of who owns what so there is no confusion later. </context> <inputs> - Notes/transcript: [paste transcript or notes] - Attendees: [names and roles] </inputs> <task> Produce a clean ownership table: every deliverable, decision, and action mapped to its single owner. Where ownership was shared or unclear, surface it explicitly. </task> <constraints> - One primary owner per item; list any supporting contributors separately - Flag every item that left the meeting without a clear owner - Do not merge distinct deliverables under one row </constraints> <format> Markdown table: Item, Type (Deliverable/Decision/Action), Primary Owner, Supporting, Status. Below it, a short "Needs an owner" callout list. </format>

<context> I want a commitment tracker capturing exactly what each person promised in [meeting], in their own words. </context> <inputs> - Transcript: [paste transcript] - Today's date: [date] </inputs> <task> Extract every explicit verbal commitment ("I'll", "I will", "I can have it by"). Capture who promised it, the promise, the stated timeframe, and the verbatim quote. Distinguish firm commitments from soft intentions ("I might", "I'll try"). </task> <constraints> - Use only what was actually said; do not strengthen soft language into firm commitments - Mark vague timeframes as [no firm date] </constraints> <format> A table: Person, Commitment, Timeframe, Firmness (Firm/Soft), Quote. Then a short list of soft intentions worth converting into firm commitments. </format>

<context> I need a single clean next-steps slide to share after [meeting] with execs and the team. </context> <inputs> - Action items / notes: [paste transcript or actions] - Audience: [e.g. leadership review] - Overall goal or deadline: [details] </inputs> <task> Distill the action items into a presentation-ready next-steps slide. Show at most 5-7 grouped next steps, each with an owner and target date, plus one line on the headline goal. </task> <constraints> - Be concise and outcome-focused; no meeting minutiae - Group granular tasks into themes so the slide is not crowded - Plain, confident language </constraints> <format> Slide title, a one-line objective, a bulleted next-steps list (Step - Owner - Date), and an optional footer with the next checkpoint. </format>

<context> I am scheduling the follow-up to [meeting] and want an agenda driven by the open action items. </context> <inputs> - This meeting's action items and decisions: [paste] - Follow-up meeting length: [e.g. 30 min] - Attendees: [names] </inputs> <task> Build a timeboxed agenda for the follow-up. Lead with status checks on open actions, then unresolved decisions, then new topics. Allocate minutes per item and assign a discussion lead. </task> <constraints> - Total time must fit the stated meeting length - Put the highest-stakes or most-blocked items first - Include a 2-minute close for assigning new actions </constraints> <format> A table: Agenda Item, Lead, Minutes, Goal of Discussion. End with a one-line meeting objective at the top. </format>

<context> An action item from [meeting] is approaching or past its due date and I want to nudge the owner without sounding pushy. </context> <inputs> - The action and its owner: [item, owner] - Original due date and today's date: [dates] - Relationship and tone: [peer / report / cross-team, friendly] </inputs> <task> Draft a short, warm accountability nudge. Acknowledge they are busy, restate the specific item and date, offer help or to remove a blocker, and ask for a quick status. </task> <constraints> - No guilt-tripping or passive aggression - Under 80 words - End with one clear, easy-to-answer question </constraints> <format> A Subject line and message body. Then a slightly firmer alternate version in case the first nudge goes unanswered. </format>

<context> Each action item from [meeting] should become a well-formed ticket in [tracker, e.g. Jira/Linear]. </context> <inputs> - Action items / notes: [paste transcript or actions] - Tracker: [Jira / Linear / GitHub Issues] - Default labels or components: [list] </inputs> <task> For each action item, write a complete ticket: a clear title, a description with context, acceptance criteria, owner, priority, and suggested labels. Phrase titles as actions and acceptance criteria as testable bullet points. </task> <constraints> - Acceptance criteria must be specific and verifiable, not vague - Include enough context that someone not in the meeting could pick it up - Leave owner blank and flag if none was named </constraints> <format> One ticket per action, each as a titled block with fields: Title, Description, Acceptance Criteria, Owner, Priority, Labels. </format>

<context> There are more action items from [meeting] than the team can do at once and I need a defensible priority order. </context> <inputs> - Action items: [paste action list] - Key goal or deadline this serves: [details] - Any known constraints: [capacity, dependencies] </inputs> <task> Rank every action by priority using impact, urgency, effort, and dependency as criteria. Give each a tier (P0-P3) and a one-sentence rationale tying it to the goal. Call out quick wins (high impact, low effort) separately. </task> <constraints> - Every ranking must include explicit reasoning, not just a label - Be honest when two items are genuinely tied </constraints> <format> A ranked table: Rank, Action, Priority Tier, Impact, Effort, Rationale. Then a short "Do first" shortlist of quick wins. </format>

<context> Many action items from [meeting] have no due date and I need sensible deadlines proposed. </context> <inputs> - Action items: [paste action list] - Overall milestone or hard deadline: [date, if any] - Today's date and team cadence: [e.g. weekly sprints] </inputs> <task> For each dateless action, propose a realistic due date. Base it on effort, dependencies, the overall milestone, and the team's cadence. Explain the reasoning behind each proposed date in one line. </task> <constraints> - Clearly mark every proposed date as [proposed] versus any date that was actually stated - Work backward from the milestone so the dates are internally consistent - Do not bunch everything on the final day </constraints> <format> A table: Action, Owner, Stated Date, Proposed Date, Reasoning. End with a note on any actions that cannot fit before the milestone. </format>

<context> Too many action items from [meeting] landed on a few people and I want delegation suggestions to balance the load. </context> <inputs> - Action items with current owners: [paste action list] - Team and their strengths or availability: [names, skills, capacity] </inputs> <task> Spot owners who are overloaded and actions that could be delegated. For each, suggest a better-suited or less-loaded owner and explain why, including what context they would need to take it on. </task> <constraints> - Only suggest delegation where it genuinely fits the person's skills or capacity - Keep accountability clear; note where the original owner should still oversee - Do not delegate items that truly require the named owner </constraints> <format> A table: Action, Current Owner, Suggested Owner, Why, Handoff Notes. Then a one-line load summary showing item counts per person before and after. </format>

<context> I want one weekly digest combining action items from all my meetings this week. </context> <inputs> - Notes or action lists from each meeting: [paste multiple, labeled by meeting and date] - My name: [name] - Week ending date: [date] </inputs> <task> Merge all action items into a single rollup. De-duplicate items that appeared in more than one meeting, and group everything by status: Overdue, Due This Week, Upcoming, Done. Within each group, show owner, source meeting, and due date. </task> <constraints> - Merge duplicates and note which meetings they came from - Sort each group by due date - Surface a "needs my attention" subset for items I own </constraints> <format> Sections by status with a sub-table each, then a short top-of-mind summary of my overdue and due-this-week items. </format>

<context> We made several decisions in [meeting name] and I want a durable log so future-us remembers WHY, not just what. </context> <inputs> - Transcript or notes: [paste here] - Date and attendees: [fill in] </inputs> <task> For each decision actually made, capture: Decision, Rationale (the real why), Alternatives considered, Owner, and a confidence read. </task> <constraints> - Only include decisions that were truly settled; list near-decisions separately under [Pending]. - Quote the meeting where it sharpens the rationale; otherwise paraphrase tightly. </constraints> <format> Markdown artifact, one section per decision, newest first. </format>

<context> In [meeting name] we picked one path among several. I want the chosen option AND the discarded ones documented so nobody relitigates settled ground. </context> <inputs> - Notes or transcript: [paste] - The decision in question: [e.g. which vendor / which architecture / which launch date] </inputs> <task> 1. List every option that was genuinely on the table. 2. For each, summarize the case for and the case against as discussed. 3. State which option won and the decisive reason it beat the runner-up. </task> <constraints> - Represent rejected options fairly, not as strawmen. - If the deciding reason was implicit, infer it and mark it [inferred]. </constraints> <format> Short intro line, then one block per option, then a "Why this one won" closing paragraph. </format>

<context> Not every decision deserves the same scrutiny. I want each decision from [meeting name] tagged by reversibility so we slow down only where it matters. </context> <inputs> - Decisions or notes: [paste] </inputs> <task> For each decision, classify it as: - One-way door (hard or costly to reverse), or - Two-way door (cheap to undo) Then give a one-line justification and the rough cost or effort to reverse. </task> <constraints> - When reversibility is ambiguous, default to one-way and explain why. - Flag any two-way-door decision we are treating with one-way-door caution (and vice versa) as a [process note]. </constraints> <format> Table: Decision | Door type | Cost to reverse | Justification. Add a short note flagging mismatches. </format>

<context> Decisions from [meeting name] need clear ownership and a stakeholder map so the right people are looped in and accountable. </context> <inputs> - Transcript or notes: [paste] - Known roles or org context: [optional, fill in] </inputs> <task> For each decision, identify: - The decision-maker (who actually owns the call) - Who was consulted - Who must be informed - Anyone who should have been in the room but was not </task> <constraints> - Use names where stated, roles where not. - Do not invent stakeholders; mark gaps as [unassigned - needs confirming]. </constraints> <format> RACI-style table per decision: Decision | Responsible | Consulted | Informed | Missing voices. </format>

<context> Decisions rest on assumptions that often go unspoken. From [meeting name] I want those assumptions made explicit so we can pressure-test them later. </context> <inputs> - Notes or transcript: [paste] </inputs> <task> For each decision, list the assumptions it depends on (about the market, users, budget, timeline, or another team). For each assumption, note: - How confident the room seemed - How we would know if it turned out false </task> <constraints> - Separate load-bearing assumptions (decision collapses if wrong) from minor ones. - Phrase each assumption as a falsifiable statement, not a vague hope. </constraints> <format> Per decision: a bulleted list of assumptions, each tagged [load-bearing] or [minor]. </format>

<context> A significant decision was made in [meeting name] and I want a formal ADR so it lives in the repo and survives team turnover. </context> <inputs> - Discussion notes or transcript: [paste] - Decision title: [e.g. Adopt event-driven messaging for order service] </inputs> <task> Draft a complete Architecture Decision Record with these sections: Status, Context, Decision, Consequences (positive and negative), and Alternatives Considered. </task> <constraints> - Keep Context factual and free of advocacy; reserve the recommendation for Decision. - State consequences honestly, including the ones we would rather not admit. - Default Status to "Accepted" unless the notes say otherwise. </constraints> <format> Markdown artifact following the standard ADR template, ready to commit as a .md file. </format>

<context> The most useful decisions are sometimes the things we ruled out. From [meeting name] I want a clear record of explicit non-decisions and de-scoping calls. </context> <inputs> - Notes or transcript: [paste] </inputs> <task> Capture everything the team explicitly chose NOT to pursue: features cut, ideas parked, approaches rejected, scope trimmed. For each, note the reason and whether it is "no for now" or "no forever". </task> <constraints> - Include only deliberate rejections, not topics that simply went undiscussed. - Distinguish a true rejection from a deferral and label each accordingly. </constraints> <format> Two grouped lists: "Decided against (permanent)" and "Parked (revisit later)", each with one-line reasons. </format>

<context> Not everything got settled in [meeting name]. I want the loose ends tracked so they do not silently become someone problem at the worst moment. </context> <inputs> - Transcript or notes: [paste] </inputs> <task> Extract every open question, unresolved tension, or "we will figure that out later" moment. For each, capture: - The question - Why it is unresolved (missing data, missing person, no consensus) - Who is best placed to resolve it - How urgent it is </task> <constraints> - Phrase each as a crisp question, not a vague worry. - Do not pad the list with rhetorical questions that were already answered. </constraints> <format> Table: Open question | Blocked on | Owner | Urgency (now/soon/later). </format>

<context> We made a decision in [meeting name] that the wider team needs to hear about clearly and without drama. </context> <inputs> - The decision and its rationale: [paste notes] - Audience: [e.g. whole engineering org / the company / one squad] - Channel: [e.g. Slack post / email / wiki update] </inputs> <task> Write an internal announcement that states what was decided, why, what changes for the reader, and what they should do next. </task> <constraints> - Lead with the decision, not the backstory. - Plain, confident tone; no corporate filler. - Pre-empt the two most likely objections in a short FAQ. </constraints> <format> Headline, two to four short paragraphs, a "What this means for you" line, and a mini FAQ. </format>

<context> Every decision from [meeting name] carries risk. I want those risks named up front rather than discovered in the post-mortem. </context> <inputs> - Decisions or notes: [paste] </inputs> <task> For each decision, identify the main risks it introduces. For each risk, estimate likelihood and impact, and suggest one mitigation or early-warning signal. </task> <constraints> - Focus on risks the decision creates, not generic project risks. - Rate likelihood and impact as low/medium/high; no false precision. - Flag any risk with no owner as [unowned]. </constraints> <format> Per decision: a table of Risk | Likelihood | Impact | Mitigation / early signal. </format>

<context> The decisions from [meeting name] are not independent; some only hold if others do. I want the dependency web made visible. </context> <inputs> - Decisions or notes: [paste] - Related prior decisions or constraints: [optional] </inputs> <task> 1. List the decisions. 2. For each, identify what it depends on and what depends on it. 3. Flag any decision that is blocked, any circular dependency, and the right sequence to execute them. </task> <constraints> - Include dependencies on other teams, tools, or external events, not just internal calls. - Call out any decision made out of order (committed before its prerequisite). </constraints> <format> A dependency list per decision, then a suggested execution order, then a short "blocked or out-of-sequence" callout. </format>

<context> Some decisions from [meeting name] are right for now but should be re-examined once conditions change. I want explicit revisit triggers so they do not calcify. </context> <inputs> - Decisions or notes: [paste] - Today date and any known milestones: [fill in] </inputs> <task> For each decision, recommend a "revisit by" date OR a triggering event (whichever fits better). Explain what new information by then should prompt a fresh look. </task> <constraints> - Prefer event triggers over arbitrary dates where a clear event exists. - Decisions that are genuinely permanent should be marked [no revisit needed] with a reason. </constraints> <format> Table: Decision | Revisit by (date or trigger) | What to re-check | Why then. </format>

<context> We committed to [decision] in [meeting name]. Before we execute, I want a pre-mortem: imagine it is six months out and this decision failed. </context> <inputs> - The decision and its rationale: [paste] - Relevant context or constraints: [optional] </inputs> <task> 1. Imagine the decision has clearly failed. Write the plausible story of how it went wrong. 2. List the top failure modes ranked by likelihood. 3. For the top three, propose a guardrail we could put in place now. </task> <constraints> - Be specific and uncomfortable; vague risks are useless here. - Separate failures we can prevent from failures we can only detect early. </constraints> <format> A short failure narrative, a ranked failure-mode list, then a "guardrails to add now" section. </format>

<context> I want an honest critique of HOW we made [decision] in [meeting name], not just whether the outcome looks good. </context> <inputs> - Transcript or notes: [paste] - The decision reached: [fill in] </inputs> <task> Assess the decision-making process: Was the problem framed well? Were enough options generated? Was evidence used or just opinion? Were biases (sunk cost, anchoring, groupthink, HiPPO) visible? Then state what would have made the decision stronger. </task> <constraints> - Critique the process and reasoning, not the people. - Be candid; do not soften real weaknesses into compliments. </constraints> <format> Scorecard across the dimensions above (strong / adequate / weak) plus a prioritized "what would strengthen it" list. </format>

<context> [Meeting name] had real disagreement before we landed on [decision]. I want the dissent recorded fairly so the minority view is not erased. </context> <inputs> - Transcript or notes: [paste] </inputs> <task> 1. Lay out each distinct viewpoint and its strongest argument. 2. Note where the disagreement actually came from (different facts, values, risk tolerance, or priorities). 3. State what was decided and what the dissenters would want re-examined if things go sideways. </task> <constraints> - Steel-man every position; give the losing side its best case. - Do not reveal who held which view unless it is essential. </constraints> <format> One section per viewpoint, a "root of the disagreement" note, then a "what dissenters want watched" closing. </format>

<context> Key people missed [meeting name] where we decided [decision]. I want a self-contained memo that brings them fully up to speed without a re-run meeting. </context> <inputs> - Notes or transcript: [paste] - Who the memo is for: [e.g. the VP / a partner team / a delayed exec] </inputs> <task> Write a rationale memo: the decision, the context that forced it, the options weighed, why this one, and what we need from the reader (if anything). </task> <constraints> - Assume the reader has zero meeting context; explain just enough, no transcript dump. - Anticipate their likely concern given their role and address it directly. - Make it skimmable in under two minutes. </constraints> <format> Memo with a one-line summary at top, then Context / Decision / Why / What we need from you. </format>

<context> A decision in [meeting name] should become a standing rule, not a one-off, so the team applies it consistently going forward. </context> <inputs> - The decision and its reasoning: [paste] - Who the policy governs: [e.g. all PRs / all vendor purchases over a threshold] </inputs> <task> Turn the decision into a clear policy or guideline doc: the rule, the rationale, when it applies, exceptions and who can grant them, and a worked example of following it. </task> <constraints> - Write the rule as an unambiguous, testable statement. - Spell out the edge cases people will actually argue about. - Keep it short enough that people will read it. </constraints> <format> Policy doc: Rule (one sentence) / Why / Scope / Exceptions / Example. </format>

<context> [Meeting name] weighed several options for [decision]. I want a clean trade-off table that captures how they stacked up across the criteria that mattered. </context> <inputs> - Notes or transcript: [paste] - The options and the criteria discussed: [fill in, or let Claude infer] </inputs> <task> 1. Identify the options and the decision criteria (cost, speed, risk, maintainability, etc). 2. Rate each option on each criterion based on the discussion. 3. Note which criteria carried the most weight and why the chosen option fit them. </task> <constraints> - Use a consistent rating scale and define it. - Mark any cell where the discussion gave no signal as [not assessed], do not guess. </constraints> <format> A comparison table (options as rows, criteria as columns) plus a short "what tipped it" paragraph. </format>

<context> For [decision] made in [meeting name], I want an honest confidence read and a clear statement of what new information would flip it. </context> <inputs> - The decision and discussion: [paste] </inputs> <task> 1. Rate confidence in the decision (low / medium / high) and justify it. 2. List the specific evidence that would raise confidence. 3. List the specific evidence that would force a reversal. 4. Recommend the cheapest test or data point to gather next. </task> <constraints> - Tie the confidence rating to evidence quality, not gut feel. - The "would change our mind" items must be concrete and observable. </constraints> <format> Confidence verdict with reasoning, then "would raise" / "would reverse" lists, then a single recommended next probe. </format>

<context> I have a long, rambling transcript from [meeting name] and I cannot tell what we actually DECIDED versus what we merely talked about. </context> <inputs> - Full transcript: [paste] </inputs> <task> Separate the signal from the noise into three buckets: 1. Decisions made (settled, with an owner if stated) 2. Action items (someone will do something) 3. Discussion only (explored but not concluded) Anything that sounds like a decision but lacks commitment goes in bucket 3 with a note on what is missing. </task> <constraints> - Be strict: a decision requires a clear commitment, not just agreement in tone. - Do not let lengthy discussion inflate into a phantom decision. </constraints> <format> Three clearly labeled lists, each item one line, with owners where stated. </format>

<context> I run or attend a daily standup for [team name]. I want a clean summary I can save or share. </context> <inputs> - Raw standup notes or transcript: [paste] - Date: [e.g. Tue 10 Jun] </inputs> <task> Summarize the standup into three sections per person: Yesterday (done), Today (planned), Blockers (and who can help). Then add a one-line "Team pulse" noting overall momentum. </task> <constraints> - Keep each person to 1-2 bullets per section - Surface anything that sounds at risk of slipping - Do not invent items not in the notes </constraints> <format> Markdown grouped by person, then the Team pulse line at the bottom. </format>

<context> I'm a [manager/IC] and just finished a 1:1 with [name]. I want structured notes plus a clear action list. </context> <inputs> - Raw 1:1 notes: [paste] - My role vs theirs: [e.g. I'm their manager] </inputs> <task> Produce two parts: 1. A short summary of what we discussed, grouped by theme. 2. An action items table with columns: Action, Owner (me or them), Due, Why it matters. </task> <constraints> - Only list actions that were actually agreed or strongly implied - Mark any item where the owner is ambiguous with "(confirm owner)" - Keep the summary under 120 words </constraints> <format> Summary as bullets, then a Markdown table for actions. </format>

<context> I have a 1:1 with [name] coming up on [date]. I want to walk in prepared using our last conversation. </context> <inputs> - Notes from our last 1:1: [paste] - Anything new since then: [paste or "nothing"] - My goal for this 1:1: [e.g. unblock their project, check on morale] </inputs> <task> Draft talking points: follow-ups on open items from last time, new topics to raise, one growth/career question, and one open-ended check-in question. Order them so the most time-sensitive comes first. </task> <constraints> - Phrase each point as something I can actually say out loud - Flag any item I committed to last time but may not have done - Max 7 talking points </constraints> <format> A numbered agenda I can glance at during the call. </format>

<context> I manage [name] and want to keep a running development log so I can coach them and write a fair review later. </context> <inputs> - Recent observations, wins, and stumbles: [paste] - Their current role and level: [e.g. mid-level engineer] - Growth areas we've discussed: [list or "none yet"] </inputs> <task> Organize these into a development snapshot: Strengths demonstrated, Skills to grow, Stretch opportunities to offer, and one suggested coaching focus for the next month. Tie each point to a specific example from the inputs. </task> <constraints> - Be specific and evidence-based, never generic praise - Keep it constructive and forward-looking - Note where I lack enough evidence to judge </constraints> <format> Four short sections with bulleted, example-backed points. </format>

<context> Today's standup surfaced several blockers and I need to decide what to act on first. </context> <inputs> - Standup notes with blockers: [paste] - Who I can pull in to help: [names/roles] - Today's top priority: [e.g. ship the release] </inputs> <task> List every blocker, then for each assign: severity (high/med/low), who can unblock it, the smallest next step, and whether it needs a decision from me or someone above me. Rank the list by what most threatens today's priority. </task> <constraints> - Separate true blockers from mere slowdowns - Flag anything blocked on a person outside the team - If two blockers share a root cause, group them </constraints> <format> A ranked table: Blocker, Severity, Unblocker, Next step, Needs decision? </format>

<context> I'm a [role] on [team]. We did a live standup and I want a clean async update to post. </context> <inputs> - Standup notes/transcript: [paste] - Audience: [e.g. cross-functional channel] </inputs> <task> Write a concise async update grouped by Shipped, In Progress, Blocked (with who can unblock). Keep it skimmable. </task> <constraints> - Lead with blockers that need a decision - No more than 8 bullets total </constraints> <format> Slack-ready Markdown, short bullets. </format>

<context> The sprint just ended. I have all the daily standup notes and need one status update for stakeholders. </context> <inputs> - All standup notes from the sprint: [paste each day, or dump together] - Sprint goal: [e.g. ship checkout v2] - Audience: [e.g. product lead and design] </inputs> <task> Synthesize the sprint into: Goal vs outcome, What shipped, What slipped and why, Blockers that recurred, and Carryover into next sprint. Distinguish a one-day hiccup from a multi-day pattern. </task> <constraints> - Collapse repetition across days into single statements - Be honest about misses without assigning blame - Keep the whole thing to one screen </constraints> <format> Five labeled sections with tight bullets. </format>

<context> I'm a [senior manager/director] and just held a skip-level with [report's report, name or anonymized]. I want notes I can act on without breaking trust. </context> <inputs> - Raw skip-level notes: [paste] - What prompted the skip-level: [e.g. routine, or a concern] </inputs> <task> Summarize into: Themes I heard, Signals about the team and their manager, Things to act on directly, and Things to feed back to their manager carefully. Separate facts from the person's interpretation. </task> <constraints> - Protect attribution: flag anything I should not repeat verbatim - Do not over-index on a single comment - Highlight where I should follow up to verify a claim </constraints> <format> Four sections, with a short "Handle with care" note at the end. </format>

<context> I want to gauge how my team is actually feeling, using the tone and content of recent standups rather than asking directly. </context> <inputs> - Standup notes from the last [number] days: [paste] - Anything I already know is going on: [context or "nothing"] </inputs> <task> Read the notes for morale and energy signals: enthusiasm, frustration, overload, disengagement, or friction between people. Cite the specific phrasing that triggered each read. End with two gentle ways I could check in without making anyone feel watched. </task> <constraints> - Be careful: tone in standups is noisy, so rate confidence low/med/high - Do not diagnose individuals as a problem - Separate workload signals from interpersonal ones </constraints> <format> Signals list with quoted evidence and confidence, then two check-in suggestions. </format>

<context> The same problems seem to keep coming up in standup and I want to prove the pattern so I can fix the root cause. </context> <inputs> - Standup notes across several days/weeks: [paste] - Team and what they're working on: [context] </inputs> <task> Find blockers or themes that recur across more than one standup. For each, show how many times it appeared, paraphrased examples, the likely root cause, and a structural fix that would stop it returning. Rank by frequency and cost. </task> <constraints> - Only flag a pattern if it appears at least twice - Distinguish a recurring symptom from its underlying cause - Note where the data is too thin to be sure </constraints> <format> A ranked table: Recurring blocker, Times seen, Root cause, Structural fix. </format>

<context> I just finished a 1:1 with [name] and want to send a short follow-up so nothing falls through the cracks. </context> <inputs> - 1:1 notes: [paste] - Relationship: [e.g. they report to me] - Tone: [warm/neutral/formal] </inputs> <task> Draft a brief follow-up email that thanks them, recaps the two or three most important things we agreed, lists who is doing what by when, and ends with an open door for questions. Keep it human, not corporate. </task> <constraints> - Only restate commitments actually made - Keep it under 150 words - No filler or buzzwords </constraints> <format> Subject line plus email body, ready to send. </format>

<context> I want to log review-worthy moments for [name] as they happen, so my next performance review is grounded in evidence not memory. </context> <inputs> - What happened: [paste the moment, win, or concern] - Their level and core responsibilities: [context] - Competencies we evaluate on: [list or "standard for their level"] </inputs> <task> Turn this moment into a review-ready entry: the situation, what they did, the impact, and which competency or value it maps to. Note whether it's a strength to reinforce or an area to develop, and suggest the exact phrasing I might use in a review. </task> <constraints> - Stay factual and specific, no inflation - Map to a competency only if it genuinely fits - Keep each entry to a few lines </constraints> <format> A single dated log entry with a competency tag. </format>

<context> After standup I want to post a quick recap in our [channel name] Slack channel so people who missed it stay in the loop. </context> <inputs> - Standup notes: [paste] - Channel vibe: [e.g. casual eng channel] - Anything to spotlight: [a win, a shoutout, or "nothing"] </inputs> <task> Write a short Slack recap: one-line headline, the top three things in flight, any blocker that needs help, and an optional shoutout. Make it scannable and friendly, with emoji used sparingly where natural. </task> <constraints> - Keep it under 120 words - One blocker maximum, the most important one - Match the channel's casual tone without being cringe </constraints> <format> Slack-ready Markdown with a bold headline line. </format>

<context> Part of my 1:1 with [name] was an OKR/goals check-in and I want to record progress cleanly. </context> <inputs> - The check-in discussion: [paste] - Their current goals/OKRs: [list with targets] - Time left in the cycle: [e.g. 4 weeks] </inputs> <task> For each goal, capture: status (on track / at risk / off track), latest progress, what's driving or blocking it, and the next concrete step. Flag any goal that needs to be re-scoped given the time left. </task> <constraints> - Base status on evidence from the discussion, not optimism - Call out goals that haven't moved since last check-in - Suggest a re-scope only when the math clearly demands it </constraints> <format> A table per goal: Goal, Status, Progress, Blocker, Next step. </format>

<context> I want to give [name] balanced feedback and I have rough notes on what went well and what didn't. I want it structured so it lands. </context> <inputs> - My rough feedback notes: [paste] - The relationship and context: [e.g. peer, or my report after a project] - How they tend to receive feedback: [direct/sensitive/unknown] </inputs> <task> Organize the feedback using a clear structure: specific praise tied to impact, then constructive points framed as situation-behavior-impact with a suggested change. Keep praise and critique genuinely separate so neither gets diluted. End with one thing I'm asking them to try next. </task> <constraints> - Every point must reference a concrete example - No feedback sandwich that buries the real message - Stay kind but clear </constraints> <format> Two sections (What worked / What to adjust) plus one "Try next" line. </format>

<context> I have a 1:1 with [name] and several loose threads from chat, email, and past notes that I keep meaning to raise. </context> <inputs> - Open threads and to-dos: [paste anything, in any order] - Past 1:1 notes still unresolved: [paste or "none"] - How much time we have: [e.g. 30 min] </inputs> <task> Cluster the threads into a tight agenda that fits the time we have. Order it: time-sensitive first, then decisions needed, then FYIs, and reserve a slot at the end for their topics. Mark which items are quick versus need real discussion. </task> <constraints> - Drop or merge anything trivial or duplicated - Do not over-pack the agenda for the time available - Always leave space for them to bring their own items </constraints> <format> A timeboxed numbered agenda with quick/discuss tags. </format>

<context> I'm coaching [name] on [skill or goal, e.g. stakeholder communication] and want to track the coaching itself across sessions. </context> <inputs> - Today's coaching conversation: [paste] - What we're working on: [the focus] - Where they were last session: [context or "first session"] </inputs> <task> Capture: the insight or breakthrough from today, what they're committing to practice, where they got stuck, and what I should prompt or revisit next session. Note progress relative to last time so I can see the trajectory. </task> <constraints> - Keep the focus on their growth, not task delivery - Be specific about the practice commitment - Flag if we seem stuck on the same point repeatedly </constraints> <format> A dated coaching log: Insight, Practice commitment, Where stuck, Next prompt. </format>

<context> I'm a [team lead] and I need to decide which issues from recent standups are worth escalating up versus handling myself. </context> <inputs> - Recent standup notes: [paste] - Who I report to and what they care about: [context] - What I can resolve on my own: [scope] </inputs> <task> Identify the issues that genuinely warrant escalation: cross-team dependencies, resourcing gaps, risks to a committed deadline, or decisions above my authority. For each, draft a one-line escalation summary stating the ask. Separately, list what I should keep handling myself. </task> <constraints> - Escalate only what truly needs leadership input, not noise - Each escalation must state a specific ask or decision needed - Be conservative: over-escalating erodes credibility </constraints> <format> Two lists: Escalate (with the ask) and Handle myself. </format>

<context> I oversee several teams and want one leadership-level view of where things stand, built from each team's standup notes. </context> <inputs> - Standup notes per team: [paste, labeled by team] - What leadership cares about right now: [e.g. the Q3 launch] - Names of the teams: [list] </inputs> <task> Roll all teams into one digest: per-team status (green/amber/red) with a one-line reason, cross-team dependencies or collisions, the top three risks across the org, and where my attention is most needed. Stay at the altitude a leader actually reads. </task> <constraints> - One line per team, not a transcript - Surface dependencies where one team is waiting on another - Lead with red items and org-level risks </constraints> <format> A status table by team, then Cross-team risks and Where I'm needed. </format>

<context> I want to keep a reliable record of what [name] has committed to, pulled from our 1:1s and standups, so I can follow up fairly. </context> <inputs> - 1:1 and standup notes mentioning their commitments: [paste] - Any commitments I'm already tracking: [paste or "none"] - Today's date: [date] </inputs> <task> Extract every commitment they made: what they said they'd do, by when, and the source/date it was said. Compare against anything I'm already tracking, mark each as open / done / overdue / slipped, and suggest which to gently follow up on next. </task> <constraints> - Only count clear commitments, not vague intentions - Mark items overdue strictly against today's date - Keep the follow-up framing supportive, not gotcha </constraints> <format> A tracking table: Commitment, Due, Source, Status, Follow up? </format>

<context> I just finished an external call with [client/company] about [topic/project]. I need a clean recap I can send internally and act on. </context> <inputs> - Call notes/transcript: [paste] - Attendees (us / them): [fill] - Call date: [date] </inputs> <task> Produce a recap with: a 3-sentence summary, key decisions made, open questions, and a Next Steps table (action, owner, due date). Infer owners and dates only where the call makes them clear, otherwise mark [TBD]. </task> <constraints> - Keep it factual, no invented commitments - Separate "decided" from "discussed but unresolved" </constraints> <format> Markdown with a Next Steps table at the end. </format>

<context> I'm in [sales/CS] and just finished a discovery call with [prospect/company]. I need CRM-ready notes. </context> <inputs> - Call transcript/notes: [paste] - Deal stage: [fill] - Methodology we use: [BANT/MEDDIC/none] </inputs> <task> Produce: 3-line summary, pain points, current solution, budget/authority/need/timeline where known, and next steps with dates. If I named a methodology, structure the qualifying fields to match it. </task> <constraints> - Mark anything not explicitly stated as [unknown] - No embellishment, only what was said </constraints> <format> Markdown, sectioned, ready to paste into the CRM note field. </format>

<context> We ran a scoping call with [client] to define what they need built/delivered for [project]. I need a structured requirements capture. </context> <inputs> - Call notes/transcript: [paste] - Project type: [software/service/campaign/other] </inputs> <task> Extract and organize: functional requirements, non-functional requirements (performance, security, compliance), explicit out-of-scope items, assumptions, and dependencies on the client. For each requirement, tag priority as [must/should/could] based on how the client framed it. </task> <constraints> - Only capture requirements actually raised on the call - Flag any requirement that sounds ambiguous and needs clarification </constraints> <format> Markdown tables grouped by requirement type, plus an "Open clarifications" list. </format>

<context> On my call with [prospect/client] several objections came up. I want a clean log of each objection and our response for coaching and CRM. </context> <inputs> - Call transcript/notes: [paste] - Product/offer: [fill] </inputs> <task> For every objection raised, capture: the exact objection, the underlying concern behind it, how we responded, and whether it landed (resolved / partially / unresolved). Then suggest one sharper response we could have given for any objection that did not fully land. </task> <constraints> - Quote the objection in the client's own words where possible - Be honest about which objections were left hanging </constraints> <format> One block per objection, then a short "Coaching notes" section. </format>

<context> I need to send a follow-up email to [client/contact] after our call about [topic]. I want it warm, concise, and clearly moving things forward. </context> <inputs> - Call notes: [paste] - Relationship stage: [new/established] - My tone: [formal/friendly/concise] </inputs> <task> Draft a follow-up email that thanks them, recaps the 2-3 key points discussed, confirms agreed next steps, and proposes a clear single next action. Keep it skimmable on mobile. </task> <constraints> - Under 180 words - One clear call to action, no walls of text - No invented promises beyond what the notes support </constraints> <format> Subject line plus email body. Offer one shorter alternate version. </format>

<context> I have to write a proposal for [client] and want to mine our call for everything that should feed into it before I start drafting. </context> <inputs> - Call transcript/notes: [paste] - Proposal type: [statement of work/retainer/one-off project] </inputs> <task> Extract proposal-ready inputs: the client's stated goals, their success criteria, scope they described, timeline expectations, budget signals, decision process, and any sensitivities to handle delicately. Then list what is still missing before I can write a complete proposal. </task> <constraints> - Separate confirmed facts from inferred signals - Do not draft the proposal, just assemble the raw inputs </constraints> <format> Sectioned brief, ending with a "Gaps to close" checklist. </format>

<context> A scoping call with [client] gave me enough to draft a statement of work / spec. I want a structured outline I can flesh out. </context> <inputs> - Call notes/transcript: [paste] - Engagement type: [project/retainer] </inputs> <task> Build a SOW outline covering: objectives, scope of work, deliverables with acceptance criteria, milestones and timeline, assumptions, exclusions, client responsibilities, and a placeholder for commercials. Draw every item from what was actually discussed; mark gaps as [to define]. </task> <constraints> - Acceptance criteria must be concrete and testable where the call allows - Keep exclusions explicit to prevent scope creep </constraints> <format> Markdown SOW skeleton as a copyable artifact, numbered sections. </format>

<context> Multiple people from [client] were on the call. I want to map who they are, their influence, and how they feel about moving forward. </context> <inputs> - Call transcript with speaker names: [paste] - What I already know about the org: [fill] </inputs> <task> For each person who spoke, infer their likely role, their priorities, their stance (champion/neutral/skeptic), and their influence on the decision. Identify who was NOT on the call but was referenced as a decision-maker. Suggest who to nurture next and how. </task> <constraints> - Base stance on what they actually said, flag low-confidence guesses - Note any power dynamics you observed </constraints> <format> Stakeholder table plus a short "Who to engage next" section. </format>

<context> I want a clear-eyed read on the risks and warning signs that surfaced during my call with [client]. </context> <inputs> - Call notes/transcript: [paste] - Deal/project context: [fill] </inputs> <task> Identify risks and red flags across: commercial (budget, payment), delivery (scope, timeline, resourcing), relationship (trust, alignment, decision chaos), and external (compliance, dependencies). For each, rate severity [low/med/high], cite the moment that triggered it, and suggest one mitigation. </task> <constraints> - Distinguish a genuine red flag from normal early-stage uncertainty - Be candid, do not soften real warning signs </constraints> <format> Risk table sorted by severity, mitigations in the last column. </format>

<context> During the call we made commitments to [client]. I need a precise log so nothing we promised slips. </context> <inputs> - Call transcript/notes: [paste] - Who from our side was on the call: [fill] </inputs> <task> Extract every commitment WE made to the client: what was promised, who promised it, any stated deadline, and what the client now expects as a result. Separate firm commitments from soft "we'll look into it" statements. Flag anything we promised that may be hard to deliver. </task> <constraints> - Capture exact wording of each promise - Do not include client-side commitments here, only ours </constraints> <format> Commitment table plus an "At-risk promises" callout. </format>

<context> I want an honest read on how [prospect/client] actually felt during our call and how close they are to buying. </context> <inputs> - Call transcript/notes: [paste] - Where we are in the deal: [fill] </inputs> <task> Assess overall sentiment and momentum. Pull positive buying signals (urgency, budget talk, asking about onboarding) and negative signals (stalling, vague timelines, looping in more approvers). Give a deal-temperature read [hot/warm/cool] with the evidence behind it, and the single biggest thing standing between us and a yes. </task> <constraints> - Cite the specific lines that drove each signal - Resist optimism bias, weigh the cool signals fairly </constraints> <format> Sentiment summary, signals split into two columns, then a temperature verdict. </format>

<context> I just ran a QBR with [client] reviewing the last quarter and planning the next. I need a recap for the account file and the client. </context> <inputs> - QBR notes/transcript: [paste] - Key metrics presented: [fill] - Account health going in: [green/yellow/red] </inputs> <task> Produce a QBR recap: wins and value delivered last quarter, metrics vs targets, issues raised, agreed priorities for next quarter, and renewal/expansion signals. End with an updated account-health call and the top 3 actions for our team. </task> <constraints> - Tie claimed wins to evidence or metrics, not vibes - Surface any dissatisfaction the client voiced, even if minor </constraints> <format> Two artifacts: an internal account recap and a shorter client-facing summary. </format>

<context> We just held the kickoff/onboarding call with new client [client]. I want a checklist that captures everything needed to onboard them cleanly. </context> <inputs> - Onboarding call notes: [paste] - What we're delivering: [product/service] </inputs> <task> Build an onboarding checklist from the call: access/credentials needed, info still to collect from the client, internal setup tasks, key contacts and their roles, agreed timeline and milestones, and success criteria they defined. Assign each item an owner [us/client] and flag blockers that hold up the start. </task> <constraints> - Mark any unconfirmed item as [confirm with client] - Order tasks so blockers appear first </constraints> <format> Checklist grouped by phase, owner tagged on every line. </format>

<context> On my call with [client] they asked for features or changes to our product. I need these logged cleanly for the product team. </context> <inputs> - Call notes/transcript: [paste] - Product area: [fill] - Account value/tier: [fill] </inputs> <task> Extract each feature request: what they asked for, the underlying job-to-be-done behind it, the use case, how blocking it is for them, and whether a workaround exists today. Reframe vague asks into a clear problem statement product can act on. </task> <constraints> - Separate the literal request from the real underlying need - Do not commit us to building anything; this is intake only </constraints> <format> One structured card per request, ready to paste into the product backlog. </format>

<context> During our call, [client/prospect] mentioned competitors or alternatives they are considering or using. I want this captured for sales and competitive intel. </context> <inputs> - Call transcript/notes: [paste] - Our product: [fill] </inputs> <task> Extract every competitor or alternative mentioned, what the client said about each (likes, gripes, pricing, gaps), where we stack up by their own words, and any switching triggers they hinted at. Flag intel that is fresh or surprising versus what we already assume. </task> <constraints> - Attribute each point to the client's statement, do not add outside knowledge - Note confidence where the client was vague </constraints> <format> Per-competitor summary, then a short "Positioning angles for this deal" list. </format>

<context> We had a pricing/negotiation conversation with [client]. I need a precise summary so I can prep the counter and not misremember terms. </context> <inputs> - Call notes/transcript: [paste] - Our list pricing/terms: [fill] - Our walk-away/floor: [fill] </inputs> <task> Summarize: the client's asks (price, terms, scope), the rationale they gave, what we offered or conceded, what is still open, and the gap between us. Identify their likely priorities (price vs scope vs terms) and suggest a counter that protects margin. </task> <constraints> - Capture exact numbers and terms discussed, no rounding - Never propose going below the floor I gave you </constraints> <format> Negotiation summary plus a recommended counter with reasoning. </format>

<context> I want an early read on whether [client] is heading toward renewal or churn based on what they said on our call. </context> <inputs> - Call notes/transcript: [paste] - Renewal date: [fill] - Current account health: [green/yellow/red] </inputs> <task> Identify renewal signals (value realized, expansion interest, advocacy) and churn signals (low usage, unmet expectations, budget pressure, champion leaving). Weigh them into a renewal-risk call [low/med/high] and recommend the 2-3 highest-leverage save or grow actions before the renewal date. </task> <constraints> - Ground every signal in something the client actually said - Be willing to call high risk even on a friendly call </constraints> <format> Two signal columns, a risk verdict, then a prioritized action plan. </format>

<context> I closed [client] and need to hand the account over to Customer Success cleanly so nothing is lost between sales and delivery. </context> <inputs> - Discovery/closing call notes: [paste] - What they bought: [fill] - Promises made during the sale: [fill] </inputs> <task> Write an internal handoff note: why they bought, their goals and success criteria, stakeholders and roles, commitments made during the sale, known risks or sensitivities, and recommended onboarding priorities. Highlight anything CS must honor that the client is expecting. </task> <constraints> - Be explicit about sales promises so CS does not under-deliver - Flag any expectation that may have been oversold </constraints> <format> Internal memo with clear sections, ready to drop into the CS workspace. </format>

<context> After my call with [client] I want to send a short email that confirms in writing exactly what we agreed, so both sides are aligned. </context> <inputs> - Call notes: [paste] - Tone: [formal/friendly] </inputs> <task> Write a confirmation email that thanks them, states the agreed actions as a clear bulleted list with owner and date for each, and invites them to correct anything I misremembered. Keep it tight and unambiguous so it doubles as a record of agreement. </task> <constraints> - Only include actions actually agreed, mark dates as [proposed] if not firmly set - Plain, professional language, no filler </constraints> <format> Subject line plus email body with a bulleted action list. </format>

<context> [Client] said a lot on our call, but I want to get beneath the surface to what they are really trying to achieve and why. </context> <inputs> - Call transcript/notes: [paste] - Context on the account: [fill] </inputs> <task> Separate what they literally asked for from the deeper outcome they actually want. Identify their underlying motivations, the fears or pressures driving them, who they need to look good in front of, and the unstated success they are quietly measuring us against. Then state the single most important thing we must get right to win them. </task> <constraints> - Distinguish evidence-based reads from inferences, and label inferences as such - Avoid putting words in their mouth; tie reasoning to what they said </constraints> <format> "Said" vs "Actually wants" comparison, motivations, then a one-line "What matters most". </format>

<context> I took rough notes during a board meeting and need formal, accurate minutes for the board pack and the record. </context> <inputs> - Rough notes or transcript: [paste] - Meeting date and location: [fill] - Attendees, apologies, guests: [fill] - Quorum requirement: [e.g. majority of directors] </inputs> <task> Produce formal minutes covering: attendance and quorum, approval of prior minutes, each agenda item with a Discussion / Decision / Action split, and a numbered resolutions list at the end. </task> <constraints> - Neutral, factual, third-person past tense - Keep resolved decisions separate from discussion - Do not invent attendees, figures, or outcomes - Mark anything ambiguous as [to confirm] </constraints> <format> Markdown artifact formatted for the board pack, with a header block and a resolutions appendix. </format>

<context> A leadership meeting covered many topics but I only need a clean record of the decisions that were actually made. </context> <inputs> - Notes or transcript: [paste] - Decision-makers present: [fill] </inputs> <task> Extract every decision that was made (not merely discussed). For each, capture: the decision, who made or owns it, the rationale given, the effective date, and any conditions attached. </task> <constraints> - Include only firm decisions, not open discussion or parked items - Flag any decision where the owner is unclear as [owner TBD] - Distinguish reversible from one-way-door calls where stated </constraints> <format> A decisions table plus a one-paragraph executive summary above it. </format>

<context> Board members made asks of the executive team, and the executive team committed to things in return. I want to track both sides. </context> <inputs> - Meeting notes or transcript: [paste] - Prior open commitments still outstanding: [paste or "none"] </inputs> <task> Build a two-way tracker. Capture (a) every ask the board made of management and (b) every commitment management made to the board, each with owner, due date, and current status. </task> <constraints> - Separate board-to-management asks from management-to-board commitments - Carry forward any prior open items and mark them [carried over] - Use status values: Not started, In progress, Done, Blocked </constraints> <format> Two tables (Board asks, Management commitments) plus a short list of anything overdue. </format>

<context> The board discussion contains raw material I need for the monthly written investor update, but it is buried in conversation. </context> <inputs> - Board discussion notes or transcript: [paste] - Investor update sections I usually send: [e.g. highlights, lowlights, metrics, asks, runway] </inputs> <task> Mine the discussion for content that maps to each investor update section. Pull out quotable highlights, candid lowlights, any metrics mentioned, and specific asks the company could make of investors. </task> <constraints> - Only use facts and figures actually stated, never estimate - Keep lowlights honest, not spun - Flag numbers that need verification as [verify] - Do not include anything marked confidential to the board </constraints> <format> Draft investor update sections, each as a short bulleted block I can edit. </format>

<context> The leadership team debated many things, but I need the short list of strategic priorities they actually aligned on. </context> <inputs> - Meeting notes or transcript: [paste] - Number of priorities to land on: [e.g. 3 to 5] </inputs> <task> Distill the discussion into the agreed strategic priorities for the period. For each, capture the priority, why it matters now, the executive sponsor, and how success will be judged. </task> <constraints> - Cap at the number requested; force ranking if there are more - Separate agreed priorities from things merely raised - Each priority must have a sponsor; mark [unassigned] if none </constraints> <format> A ranked priorities list followed by a short "explicitly deprioritized" note. </format>

<context> We ran a multi-session leadership offsite and I have scattered notes across several whiteboards and breakout groups. </context> <inputs> - All session and breakout notes: [paste] - Offsite goals set at the start: [fill] - Sessions held: [list] </inputs> <task> Synthesize the whole offsite into one coherent readout: themes that emerged, decisions reached, open debates left unresolved, and concrete next steps. Connect threads that recurred across different sessions. </task> <constraints> - Tie the synthesis back to the stated offsite goals - Name unresolved tensions honestly rather than smoothing them over - Attribute decisions to owners where stated </constraints> <format> An offsite readout artifact: Themes, Decisions, Open questions, Next steps, with a one-paragraph opener. </format>

<context> We reviewed company OKRs at the leadership meeting and I need notes that capture where each objective stands. </context> <inputs> - OKR review notes or transcript: [paste] - Current OKRs with targets: [paste] </inputs> <task> For each objective and key result, capture the status discussed, the latest number versus target, the trend, the blocker if behind, and the agreed action. Highlight any KR that changed scope or definition during the meeting. </task> <constraints> - Use a clear status: On track, At risk, Off track, Done - Only record numbers that were actually stated - Flag re-scoped or moved goalposts explicitly </constraints> <format> A table per objective with its KRs, plus a short list of at-risk items needing escalation. </format>

<context> The meeting included several budget and spending decisions and I need an auditable log of what was approved and why. </context> <inputs> - Meeting notes or transcript: [paste] - Budget context if relevant: [e.g. quarter, total envelope] </inputs> <task> Log every budget decision: amount, what it funds, the owner, the approval status, the rationale, and the trade-off accepted (what was cut or deferred to fund it). </task> <constraints> - Record only amounts and approvals actually stated - Capture conditional approvals with their conditions - Mark any figure needing confirmation as [verify] - Note where a decision was deferred rather than approved </constraints> <format> A budget decision log table plus a short "deferred or rejected" section. </format>

<context> I am minuting a risk committee session and need a disciplined record of risks reviewed and decisions taken. </context> <inputs> - Risk discussion notes or transcript: [paste] - Risk register categories we use: [e.g. financial, operational, regulatory, cyber, reputational] </inputs> <task> For each risk discussed, capture: description, category, likelihood and impact as stated, current mitigation, owner, residual risk view, and any change in rating since last review. Note risks newly raised and risks closed. </task> <constraints> - Do not assign ratings that were not stated; mark [rating not set] - Keep the committee's risk-appetite framing where mentioned - Separate new risks, changed risks, and closed risks </constraints> <format> A risk table plus three short lists: newly raised, rating changed, closed. </format>

<context> My leadership meeting notes mix sensitive items with things the whole company can see, and I need to split them safely before sharing. </context> <inputs> - Full meeting notes or transcript: [paste] - What counts as sensitive here: [e.g. personnel, M and A, legal, unannounced financials] </inputs> <task> Sort the content into two streams: a confidential leadership-only version and a company-shareable version. Move anything touching people, legal, financials, or unannounced plans into the confidential stream and redact it from the shareable one. </task> <constraints> - When in doubt, classify as confidential - Do not paraphrase sensitive content into the shareable version - Flag borderline items as [review before sharing] </constraints> <format> Two clearly labeled sections: CONFIDENTIAL (leadership only) and SHAREABLE (all company). </format>

<context> After the board meeting I need a tight action list so nothing agreed gets dropped before the next session. </context> <inputs> - Meeting notes or transcript: [paste] - Next board meeting date: [fill] </inputs> <task> Extract every action item agreed in the meeting. For each, capture the action, the single accountable owner, the due date, and how completion will be evidenced at the next board meeting. </task> <constraints> - One accountable owner per action; split if shared - Default due dates to before the next board meeting unless stated - Flag actions with no owner as [needs owner] - Exclude vague aspirations that are not real commitments </constraints> <format> An action table sorted by due date, plus a short list of items due before next board. </format>

<context> The board touched on governance and compliance matters that must be recorded carefully for the corporate record. </context> <inputs> - Meeting notes or transcript: [paste] - Governance matters relevant to us: [e.g. policy approvals, conflicts of interest, regulatory filings, director appointments] </inputs> <task> Capture every governance and compliance item: approvals granted, conflicts of interest declared, policies adopted or amended, regulatory or filing obligations noted, and any required follow-up with a responsible party. </task> <constraints> - Record declarations and recusals exactly as stated - Do not characterize legal matters beyond what was said; mark [legal to advise] - Note any obligation with a statutory deadline distinctly </constraints> <format> A governance log grouped by type, with a separate "statutory deadlines" callout. </format>

<context> I want an honest read on where the executive team actually agreed and where they did not, beneath the polite surface. </context> <inputs> - Meeting notes or transcript: [paste] - Topics in scope: [list or "all"] </inputs> <task> For each major topic, assess whether the team reached genuine alignment, surface-level agreement, or unresolved disagreement. Quote or paraphrase the signals that reveal each, and name who held which position where clear. </task> <constraints> - Do not manufacture conflict; only flag misalignment with evidence - Distinguish "agreed" from "went quiet" or "deferred" - Stay neutral; do not take a side </constraints> <format> A table: Topic, Alignment level, Evidence, Open question, plus a short "watch list" of unresolved tensions. </format>

<context> After a leadership meeting I need to write the all-hands or company message that shares what is appropriate without leaking sensitive context. </context> <inputs> - Leadership meeting notes: [paste] - Audience and channel: [e.g. all-hands, Slack post, email] - Tone we use internally: [e.g. direct, warm, candid] </inputs> <task> Draft a company-facing update that conveys the decisions and direction employees need, framed in our voice. Translate leadership shorthand into plain language and lead with what changes for the team. </task> <constraints> - Exclude anything confidential: personnel, legal, financials, unannounced plans - No corporate filler; say what is actually different - Acknowledge open questions honestly rather than over-promising </constraints> <format> A ready-to-post message with a clear opening line and a short "what this means for you" section. </format>

<context> The team committed to or rejected some big strategic bets and I want the reasoning captured before memory fades. </context> <inputs> - Meeting notes or transcript: [paste] - Bets or initiatives discussed: [list or "extract from notes"] </inputs> <task> For each strategic bet, document: the decision (pursue, reject, or defer), the thesis behind it, the key assumptions it rests on, what would prove it wrong, the owner, and the resources committed. </task> <constraints> - Capture the disconfirming evidence the team would watch for - Separate firm bets from exploratory ones - Record assumptions as assumptions, not facts </constraints> <format> One card per bet: Decision, Thesis, Assumptions, Kill criteria, Owner, Resources. </format>

<context> The meeting reallocated people, money, or focus across initiatives and I need a clear record of who got what and at whose expense. </context> <inputs> - Meeting notes or transcript: [paste] - Resources in play: [e.g. headcount, budget, engineering capacity, exec attention] </inputs> <task> Capture each allocation decision: what resource moved, from where to where, how much, the rationale, the owner, and what is now under-resourced as a result. </task> <constraints> - Make the trade-off explicit: every gain has a corresponding give - Record only amounts actually stated; mark estimates as [approx] - Note decisions deferred for lack of resource </constraints> <format> A from-to allocation table plus a short "now under-resourced" watch list. </format>

<context> A contentious topic was debated at length with strong views on multiple sides and I need a fair, balanced synthesis. </context> <inputs> - Debate notes or transcript: [paste] - The question under debate: [fill] </inputs> <task> Synthesize the debate fairly. Lay out each distinct position, its strongest argument, the evidence cited for it, and the concerns it raised. Then identify the genuine points of disagreement versus where people were talking past each other. </task> <constraints> - Steelman every position; do not favor the loudest voice - Attribute positions to people only where clearly stated - Separate factual disagreements from values or priority disagreements - Do not declare a winner unless one was actually agreed </constraints> <format> Position-by-position breakdown, then a "crux of the disagreement" summary and any decision reached. </format>

<context> We held a QBR with the leadership team and I need structured notes that connect performance to decisions and next-quarter plans. </context> <inputs> - QBR notes or transcript: [paste] - Functions or business units reviewed: [list] - Headline metrics presented: [paste or "extract"] </inputs> <task> Structure the QBR: per function, capture what worked, what missed, the why, the metric versus plan, and the commitment for next quarter. Then roll up cross-cutting themes and the top company-level priorities that emerged. </task> <constraints> - Tie every "what missed" to a stated cause, not blame - Use only metrics actually presented; flag gaps as [no data] - Separate retrospective from forward commitments </constraints> <format> A section per function plus a company-level rollup of themes and next-quarter priorities. </format>

<context> I am building the agenda for the next board meeting and want it grounded in the last meeting's minutes and open items. </context> <inputs> - Prior board minutes: [paste] - Open action and commitment tracker: [paste or "none"] - New topics management wants to raise: [list] </inputs> <task> Draft the next board agenda. Carry forward unresolved items from prior minutes, slot in follow-ups on past decisions, add the new topics, and suggest a time allocation and an owner for each item. Flag anything overdue for board attention. </task> <constraints> - Every carried-over item must trace to a prior decision or action - Order items by importance, not chronology - Mark items needing a board vote distinctly - Keep the agenda realistic for the meeting length </constraints> <format> A board agenda with item, owner, time, type (inform / discuss / decide), plus an "overdue from last time" callout. </format>

<context> The leadership meeting included sensitive people and org decisions that must be recorded discreetly and kept tightly held. </context> <inputs> - Meeting notes or transcript: [paste] - Who may see this record: [e.g. CEO, head of people only] </inputs> <task> Capture each people or org decision: the change (hire, exit, promotion, reorg, comp), who is affected, the owner, the timing, the communication plan, and any approval still required. Keep names handled with care. </task> <constraints> - Treat this as strictly confidential throughout - Do not speculate on reasons beyond what was stated - Flag anything requiring legal or HR sign-off as [needs sign-off] - Note where an announcement is gated on other steps </constraints> <format> A confidential decision log with a header warning, grouped by type of change. </format>