prompts / claude-for-work

CLAUDE FOR WORK

Lecture-aligned copy-paste prompts for the Claude for Work curriculum — Projects, Artifacts, business workflows, and developer topics. Related hub: /r/claude-for-work. Use Print → Save as PDF for an offline copy.

PROMPT BANK

20 copy-paste prompts aligned with the course scripts. Paste into Claude (adapt names, metrics, and file uploads). Claude 4.x responds best to explicit structure — see XML examples below.

01

L3 — First win: competitive analysis brief

I'm a product manager at a fintech startup. Draft a one-page competitive analysis of the top three personal finance apps. Use a professional, data-driven tone. After the draft, add a comparison table with columns: pricing, standout features, and typical user ratings (cite public sources if you use search).
02

L4 — Project “Identity” (brand voice block)

You are the writing assistant for [Company Name]. Voice: confident, human, no filler. Audience: [ICP]. Forbidden: hype words like “revolutionary” and “game-changing.” When unsure, ask one clarifying question before writing.
03

L7 — Vague vs explicit (rewrite drill)

First, answer briefly with a vague prompt: “Write about our new product.” Then answer again using this explicit brief: “Write a ~500-word product announcement blog post for our new API monitoring feature. Audience: DevOps engineers at mid-size companies. Tone: technically confident but accessible. Include one concrete use case where a team cut downtime by ~40%. End with a CTA to start a free trial.” Compare both outputs in 3 bullets.
04

L7 — XML: SaaS Q4 analyst brief

<role>Senior financial analyst specializing in SaaS metrics</role>
<context>B2B SaaS, Series B, $15M ARR. Q4 2025 data is attached / pasted below.</context>
<task>Analyze Q4 and name the three most actionable improvements.</task>
<output_format>
1. Executive summary (3 sentences)
2. Three findings with supporting numbers
3. Recommended actions with expected impact
</output_format>
<constraints>Do not speculate beyond the data; flag uncertainty explicitly.</constraints>
05

L7 — Few-shot executive summary style

Here is an example of the executive summary style I want: [paste example]. Now write a summary in the same structure and voice for [paste Q1 metrics or narrative].
06

L7 — Long-document order (query last)

[Paste or upload the full document first, above this line.] Given everything above: extract the three decisions most likely to affect revenue in the next 90 days, with citations to section headings or quotes where possible.
07

L8 — Customer billing recovery email

Draft a response to this customer email about a billing error. Context: 3-year customer, relationship matters. Goals: acknowledge the mistake briefly, explain what happened in plain language, offer a concrete resolution, close warmly. Max ~150 words.
08

L8 — Email tone variants (A/B)

Using the draft you just wrote: create two more versions — (A) more formal for enterprise buyers, (B) warmer for SMB customers. Keep facts identical.
09

L8 — Long PDF → one-page exec brief

Read the attached quarterly report PDF. Produce a one-page executive brief: QoQ metric moves, three biggest wins, two risks, and recommended priorities for next quarter. Reference page numbers for each major claim.
10

L8 — Board-ready talking points

Turn the executive brief into a 5-minute spoken script for my board: short opener, three talking points with pauses, and one ask at the end.
11

L8 — Meeting transcript → tracker rows

Clean up this messy transcript. Output: (1) decisions, (2) action items with owner + due date, (3) open questions, (4) 3-sentence summary. Then format action items as a Markdown table I can paste into Notion or Jira.
12

L9 — CSV: first-pass trends

I uploaded a CSV of sales rows. Identify the five most important trends (with % or $ where possible). Call out seasonality, outliers, or segment shifts if the data supports it.
13

L9 — Monthly revenue by region (chart)

From the same CSV: build an interactive chart in an Artifact — monthly revenue by region with per-series trendlines. Explain one insight an exec should act on this week.
14

L9 — RFM-style scatter insight

Segment customers by purchase frequency and average order value; propose clusters; produce a labeled scatter plot in an Artifact and name the highest-value cluster with evidence.
15

L10 — Branded 8-slide storyline

Turn this executive brief into an 8-slide storyline: title, exec summary, one slide per major finding with a simple chart where useful, recommendations, appendix if needed. Brand colors: navy #1E3A5F and green #2ECC71.
16

L10 — Excel formula + error triage

Here is a formula goal: rolling3-month average revenue for column C. Propose the formula and explain how to fill down. Then diagnose the #REF! error in cell F12 from this sheet context: [describe or paste range labels].
17

L10 — 5-email enterprise launch sequence

Inside this Project’s brand context: draft a 14-day, 5-email launch sequence for a new enterprise feature. Map: announcement → use case → social proof → FAQ → final call. Include subject, preview text, and body for each.
18

L6 — Research Mode (regulated landscape)

Use Research Mode. Topic: current state of AI regulation in the European Union — AI Act timeline, enforcement, and practical impact on B2B SaaS vendors serving EU customers. Require cited sources for major claims.
19

L6 — Extended thinking: anomaly hunt

Upload this spreadsheet. With Extended Thinking enabled: find the three most significant anomalies, hypothesize likely causes, and recommend specific next actions per anomaly. Show reasoning transparency appropriate to a CFO reader.
20

Capstone — Personal Claude system (one page)

Given my role as [title] at [company], design a one-page “Claude system”: which Projects to keep, what goes in each knowledge base, when I use Sonnet vs Opus vs Haiku, which MCP connectors matter, and my weekly maintenance checklist.