40 Best Manus AI Prompts (2026)

Research the competitive landscape for [product category] in [target market/geography]. Scope: only players active as of the last 12 months; ignore discontinued products. For each competitor identify positioning, target customer, pricing model, and 2-3 differentiators. Deliverable: a research report with these sections — Market Overview, Competitor Profiles (one per player), Positioning Map (described in text), Gaps & Opportunities, Sources Appendix. Rules: cite every claim with a source URL; if a data point is not public, write "not disclosed" rather than guess; tag any estimate with a confidence level (High/Med/Low). Use Manus 1.6 Max.

Use Wide Research to profile these [100] companies: [paste list or describe the set, e.g. "Series A-C fintech startups founded after 2021 in the EU"]. Run one sub-agent per company in parallel. Build a spreadsheet with columns: Company, Website, HQ Country, Founded Year, Employee Count, Funding Stage, Total Funding, Primary Product, Target Customer, Source URL. Rules: one row per company; leave a cell blank if the fact isn't found on an official or reputable source rather than guessing; put the citation URL in the Source URL column; stop when every company has a row.

Produce a literature review on [research topic/question]. Scope: peer-reviewed papers and reputable preprints from [2021]-present; prioritize highly cited and recent work. Deliverable: a report with sections — Question & Scope, Key Findings (grouped by theme), Points of Disagreement, Methodological Notes, Open Questions, and a References table (Title, Authors, Year, Venue, URL). Rules: cite every finding to a specific paper with a link; note sample size or study type where relevant; flag any claim you could only find in a single source as Low confidence.

Do a deep-dive research report on [company name] ([website]). Cover: business model and revenue lines, funding and investors, leadership team, product portfolio, recent news from the last 12 months, customer sentiment, and known risks. Deliverable: a one-document report with a one-paragraph executive summary up top, then a section per topic, then a Sources Appendix. Rules: cite every fact with a URL; separate confirmed facts from rumor/estimate and label estimates Med/Low; if financials aren't public, say so explicitly. Use Manus 1.6 Max for the reasoning.

Research recent regulatory and policy developments affecting [industry] in [region] over the last [6] months. Scope: official regulator pages, government sites, and major trade press only — no forums or social posts. Deliverable: a briefing with sections — What Changed, Effective Dates, Who It Affects, What To Do, and a dated Source List. Rules: every item must have a date and an official-source URL; if a rule is proposed but not yet in force, label it clearly; leave "Effective Date" blank rather than inventing one.

Compare these [5] tools for [job to be done]: [Tool A, Tool B, Tool C, Tool D, Tool E]. Weight the decision: capability 40%, price 30%, integrations 20%, support 10%. Deliverable: a comparison spreadsheet with columns — Tool, Starting Price, Key Features, Integrations, Free Tier (Y/N), Best For, Weighted Score, Source URL — plus a short verdict paragraph naming the winner for a [team size / use case]. Rules: pull pricing from each official pricing page and link it; leave a cell blank if not published; note the date you checked pricing.

Build a list of [50] companies that match this ICP: industry [X], headcount [50-500], region [region], and showing signals of [hiring for role / using tool / recent funding]. Deliverable: a spreadsheet with columns — Company, Website, Industry, Headcount, HQ Location, ICP Signal (which one and evidence URL), LinkedIn URL. Rules: only include companies you can verify against a real source; put the evidence link in the ICP Signal column; leave blanks rather than guess headcount; stop when you have 50 verified rows. Use Wide Research to run these in parallel.

For each company in [paste list of companies / attach sheet], find the person most likely to own [buying decision, e.g. "marketing tooling"]. Deliverable: a spreadsheet with columns — Company, Contact Name, Job Title, LinkedIn URL, Public Email (if listed), Source URL, Confidence (High/Med/Low). Rules: use only publicly available business contact info; never fabricate an email — leave it blank if not published; set Confidence to Low if you inferred the role owner rather than confirmed it; cite where you found the person.

Use the Browser Agent to pull the exhibitor/sponsor list from [event name and URL]. For each organization, capture what they do and whether they fit our ICP: [describe ICP]. Deliverable: a spreadsheet with columns — Organization, Website, Booth/Tier (if shown), What They Do, ICP Fit (Yes/No/Maybe), Reason, Source URL. Rules: only include organizations actually listed on the event site; cite the exact page; mark ICP Fit "Maybe" and explain when you're unsure rather than forcing a Yes/No.

Find companies in [industry / region] currently hiring for [role, e.g. "RevOps" or "Head of Data"], which signals they may need [our product]. Deliverable: a spreadsheet with columns — Company, Job Title, Posted Date, Job URL, Website, Headcount, Why This Is A Signal. Rules: only include live or recently posted roles (last [30] days); link the actual job posting; leave Posted Date blank if the listing doesn't show one; do not include reposts of the same role twice.

Take my attached lead list [attach spreadsheet] and enrich each row. Add columns: Website, Industry, Headcount Range, HQ Country, Funding Stage, Recent News (one line + date), Source URL. Do not change my existing columns. Rules: only fill a new cell if you can verify it from a reputable source and cite it; leave blank rather than guess; flag any row where the company name was ambiguous or you found multiple matches, so I can resolve it. Use Wide Research to process rows in parallel.

Do a full teardown of [competitor name] ([website]) versus us, [your company]. Cover: product and feature set, pricing and packaging, positioning and messaging, target segments, go-to-market channels, strengths, and vulnerabilities we can attack. Deliverable: a report with a section per topic, a side-by-side feature table (Feature, Them, Us, Notes), and a "Where We Win / Where We Lose" summary. Rules: cite every claim about the competitor with a URL; label anything inferred from marketing copy as interpretation; leave pricing blank if not published. Use Manus 1.6 Max.

Build a pricing comparison across these [6] competitors: [list]. Deliverable: a spreadsheet with one row per plan and columns — Company, Plan Name, Monthly Price, Annual Price, Key Limits (seats/usage), Notable Features, Free Trial (Y/N), Pricing Page URL, Date Checked. Rules: read each official pricing page with the Browser Agent and link it; record prices in [currency]; if a plan is "contact sales," write that instead of a number; leave any unlisted field blank. Add a short paragraph noting where we sit in the price range.

Produce a SWOT analysis of [competitor] in the context of [market/segment]. Deliverable: four sections — Strengths, Weaknesses, Opportunities, Threats — each as bullet points, every bullet backed by evidence with a source URL, plus a closing paragraph on the two moves we should make in response. Rules: keep Strengths/Weaknesses to observable facts (product, funding, team, reviews) and Opportunities/Threats to market forces; tag speculative bullets as Low confidence; no bullet without a citation or a clearly labeled inference.

Estimate TAM, SAM, and SOM for [product] targeting [customer segment] in [geography]. Show your work: state each assumption, the number you used, and the source. Deliverable: a report with a Methodology section, a sizing table (Segment, Population/Count, Avg Spend, Total, Source), the three headline figures, and a Sensitivity note showing how the SOM changes if the key assumption moves +/-20%. Rules: cite every input; label modeled figures as estimates with Med/Low confidence; never present a derived number as a hard fact.

Map the feature and positioning gaps between us, [your company], and these [4] competitors: [list], for the [use case] segment. Deliverable: a feature-coverage matrix (rows = features, columns = each vendor, cells = Yes/Partial/No with a note), plus a short list of "white space" opportunities where no vendor is strong. Rules: derive coverage from official docs, changelogs, and pricing pages and cite each cell's source; mark "Partial" with an explanation; leave a cell blank and flag it if you couldn't verify.

Analyze the attached dataset [attach CSV/Excel]. First, summarize what's in it (columns, row count, date range, obvious data-quality issues). Then answer these questions: [Q1], [Q2], [Q3]. Deliverable: a report with a Data Summary, one section per question with the supporting chart and the number, a Data-Quality Notes section, and a short "So What" list of the top 3 takeaways. Rules: state any assumption you made (how you handled nulls, outliers, duplicates); don't drop rows silently — report what you removed and why; show the exact figures behind each chart.

From the attached data [attach file], build a KPI dashboard for [team/purpose]. Metrics to show: [metric 1], [metric 2], [metric 3], each with the current value, the trend over [time period], and a comparison to [target or prior period]. Deliverable: an interactive dashboard with a chart per metric and a one-line insight under each. Rules: define each metric's formula in a Definitions section; use only the attached data — do not invent numbers; flag any metric where the data is incomplete instead of extrapolating.

Clean the attached spreadsheet [attach file]. Tasks: standardize formats (dates, phone, country names, casing), split/merge columns as needed for [describe], remove exact and fuzzy duplicates, and flag rows with missing required fields [list required fields]. Deliverable: a cleaned spreadsheet, a separate "Changes Log" sheet listing every transformation and how many rows it touched, and a "Needs Review" sheet for ambiguous rows. Rules: never delete a row without logging it; when merging duplicates, keep the most complete record and note which fields were combined.

Run a cohort retention and churn analysis on the attached subscription data [attach file]. Group customers by signup month, compute monthly retention for each cohort, and identify where churn spikes. Deliverable: a retention cohort table (heatmap-style), a churn-rate-over-time chart, and a findings section naming the worst cohort, the likely drivers you can see in the data, and 3 hypotheses to test. Rules: state your definitions of "active" and "churned"; note the sample size per cohort; label small-cohort figures (under [30] customers) as Low confidence.

Theme the open-text responses in the attached survey export [attach file], column [name]. Group responses into themes, count how many mention each, and pull 2-3 representative verbatim quotes per theme. Deliverable: a spreadsheet with columns — Theme, Mention Count, % of Respondents, Representative Quotes, Sentiment (Pos/Neg/Neutral) — sorted by count, plus a one-paragraph summary of the top 3 themes. Rules: quote verbatim, don't paraphrase; put responses that fit no theme in an "Other" bucket rather than forcing them; report the total response count analyzed.

Create a [12]-slide investor pitch deck for [company] using Manus Slides. Audience: [seed/Series A] investors. Narrative arc: Problem, Solution, Product, Market (with sizing), Business Model, Traction, Competition, Team, Ask. Deliverable: a 12-slide deck, each slide with a clear headline takeaway and minimal body text, and a speaker-notes line per slide. Rules: use only the facts and figures I provide here — [paste your key metrics] — and mark any slide where you need a number I didn't give with a [FILL] placeholder rather than inventing it.

Write a one-page executive brief on [decision/topic] for [audience, e.g. the leadership team]. Structure: a two-sentence recommendation up top, then Background, Options (with pros/cons each), Recommendation & Rationale, and Next Steps with owners and dates [I'll fill owners]. Deliverable: a single-page document, scannable, with bold key terms and no filler. Rules: lead with the recommendation; keep it to one page; if a fact is needed that I haven't provided, flag it as an open question rather than guessing.

Draft a quarterly report for [company] covering [Q_ YYYY], using the attached metrics [attach file]. Sections: Executive Summary, KPIs vs Plan (table), Wins, Misses & Why, Product Update, Financial Summary, Risks, and Priorities for Next Quarter. Deliverable: a polished document plus a matching [8]-slide Manus Slides summary. Rules: pull all numbers from the attached file and cite the sheet/cell context; never soften a miss — state it plainly with the cause; mark any figure I haven't provided as [FILL].

Create a sales one-pager for [product] aimed at [buyer persona] in [industry]. Include: the core problem, our solution, top 3 differentiators, a proof point or two [I'll provide], typical results, and a clear call to action. Deliverable: a single-page document with a strong headline, short scannable sections, and a customer-facing tone. Rules: use only claims I can back up — mark any stat or logo I need to supply as [FILL]; avoid hype words; keep every section to two or three lines.

Turn the attached document [attach file] into a [15]-slide training deck for [audience] using Manus Slides. Break the content into logical modules, one concept per slide, with a headline takeaway, 3-4 supporting bullets, and a "check your understanding" question every few slides. Deliverable: a 15-slide deck plus a one-paragraph facilitator note per module. Rules: stay faithful to the source document — don't add facts that aren't in it; if the source is thin on a topic, flag the gap rather than filling it with generic content.

Build a responsive one-page landing site for [product]. Sections: hero (headline + subhead + CTA button), 3-feature row, social proof strip, a short FAQ (5 questions), and a footer with a signup form. Style: [clean/modern, brand color #hex, sans-serif]. Copy: [paste or say "write it from this description"]. Deliverable: a working, mobile-responsive page I can preview and download. Rules: make the CTA button prominent above the fold; use placeholder [LOGO] and [FORM ENDPOINT] where I need to plug in my own; keep load-light — no heavy libraries.

Build a working prototype web app: [describe, e.g. "a ROI calculator where a user enters team size, hours saved, and hourly cost, and sees monthly and annual savings"]. Deliverable: a functioning app I can interact with in the browser, with clean inputs, live-updating results, and a mobile layout. Rules: validate inputs and handle empty/invalid entries gracefully; show the formula used somewhere on the page; keep all logic client-side; leave a clearly marked spot for me to add branding.

Use the Browser Agent to complete [task, e.g. "submit our product to these 10 startup directories"]. For each site in [paste list of URLs], fill the submission form with the details I've provided [paste company name, URL, description, category, contact email]. Deliverable: a tracking spreadsheet with columns — Site, URL, Status (Submitted/Blocked/Needs Login), Confirmation/Screenshot, Notes. Rules: never create accounts or agree to paid plans without asking me first; if a site requires login or payment, mark it "Needs Login" and skip; screenshot each confirmation page.

Use the Browser Agent to collect current prices for these [20] products from [retailer site(s)]: [paste product list or URLs]. Deliverable: a spreadsheet with columns — Product, URL, Price, Currency, In Stock (Y/N), Date/Time Checked, Notes. Rules: read the live product page for each item and link it; record the price exactly as shown including any strike-through and sale price in separate columns; leave blank and note if a page is unreachable or the price is hidden behind login; do not guess prices from cache.

Build a simple internal dashboard web app that reads the attached data [attach file] and lets me filter [dimension, e.g. by region and month] and see [key charts/tables]. Deliverable: a working app with filter controls, 3-4 visualizations that update on filter change, and a data table below. Rules: keep it read-only (no editing of source data); handle empty filter results gracefully with a friendly message; note anywhere the underlying data is incomplete; make it usable on a laptop and a tablet.

Build a 3-month SEO content plan for [website] targeting [audience] in [niche]. Research the topics and keywords real buyers search, cluster them into pillar-and-supporting groups, and prioritize by intent and difficulty. Deliverable: a spreadsheet with columns — Cluster, Working Title, Target Keyword, Search Intent, Funnel Stage, Priority (1-3), Internal Links, Notes — plus a one-paragraph strategy summary. Rules: base topics on real search behavior and cite where you saw the demand signal; don't invent search volumes — if you can't verify a number, label the priority qualitatively instead.

Build a launch campaign for [product/offer] targeting [audience]. Goal: [e.g. drive signups over 4 weeks]. Channels: email (3-email sequence), LinkedIn (5 posts), and a landing-page headline set (3 options). Deliverable: a document organized by channel with every asset written out, plus a simple 4-week calendar (Week, Channel, Asset, Status). Rules: keep one consistent core message across channels; match our tone [paste 2-3 lines of brand voice or point to Knowledge]; mark any claim needing a real stat as [FILL]; no hype words.

Audit the content marketing of these [3] competitors: [list]. For each, review their blog, key landing pages, and any resources, and identify what topics they cover, their formats, their apparent keywords, and content gaps we could own. Deliverable: a report with a per-competitor summary, a combined "topics they cover vs we cover" table, and a ranked list of 10 content opportunities for us. Rules: cite the specific pages you reviewed; base "gaps" on what you actually observed, not assumptions; tag each opportunity with effort (Low/Med/High).

From the attached article/announcement [attach or paste], produce a repurposing pack. Deliverable: one newsletter issue (subject line + 200-word body + CTA), 3 LinkedIn posts (different angles), 5 short X/social posts, and 3 subject-line variants for A/B testing. Rules: keep the core message consistent; match our voice [paste or reference Knowledge]; don't add facts not in the source; keep each social post standalone and under [280] characters where the platform requires it; no emojis unless I ask.

Write [10] ad copy variants for [product] on [platform, e.g. Google Search or Meta]. Angles to cover: [problem-led], [benefit-led], [social-proof], [urgency]. Deliverable: a spreadsheet with columns — Variant #, Angle, Headline, Primary Text, CTA, Character Count — respecting [platform]'s length limits. Rules: stay within the character limits and flag any that exceed; use only claims I can substantiate — mark unverifiable stats as [FILL]; keep at least one plain, no-hype control variant for benchmarking.

Plan a [5]-day trip to [destination] for [number] people, [dates], budget [amount] total. Preferences: [pace, interests, dietary, mobility]. Deliverable: a day-by-day itinerary document with morning/afternoon/evening blocks, specific place names with why-recommended, rough costs, a booking checklist, and a map-link column. Rules: keep the total within budget and show a running cost estimate; cite hours/prices with links and note they can change; leave a booking blank rather than guessing availability; flag anything time-sensitive I should book now.

Plan a [team offsite] for [N] people in [city], [dates], budget [amount]. Cover: venue options (3, with pros/cons and cost), a sample agenda, catering options, one team activity, and a logistics checklist. Deliverable: a planning document with a comparison table for venues (Name, Capacity, Day Rate, Distance, Link, Notes) and a timeline of what to book when. Rules: cite venue prices and capacities with links; keep the plan within budget with a cost breakdown; mark availability as "to confirm" rather than assuming; note deposit deadlines.

Set up a Scheduled Task to run every weekday at [8am my time]. Each run: check these sources [list newsletters/sites/feeds] for anything new in the last 24 hours relevant to [my role/topics], and produce a short brief. Deliverable each day: a one-page digest with sections — Top 3 Things To Know, Worth A Skim (with links), and Nothing-Burger (ignore). Rules: link every item to its source; deduplicate stories covered by multiple outlets; if nothing meaningful happened, say so in one line rather than padding.

Set up a Scheduled Task to run every [Monday at 9am]. Each week, use the Browser Agent to check these competitors [list sites] for changes since last week: pricing, new features/pages, blog posts, and job postings. Deliverable each week: a spreadsheet appended with a dated section per competitor — What Changed, Evidence URL, Why It Matters — plus a 3-bullet "This Week's Signals" summary at top. Rules: only report actual changes vs the prior run; link every change; leave a competitor's row empty and say "no change detected" rather than inventing activity.