Blueflame AI assistant and MCP prompt library

Use these prompts to see how the Blueflame AI assistant and Datasite MCP can support deal teams across the full lifecycle of a transaction, from data room setup and launch readiness to active diligence, buyer review, and final bid preparation.

 

See the Blueflame AI assistant in action

This short video brings five sample prompts to life, showing how the Blueflame AI assistant can turn common deal questions into faster, more informed action. 

Watch to see the experience in motion, then use the library below as a practical resource to explore, adapt, and combine prompts for a wider range of workflows across the transaction lifecycle. 

 

 

Note: Prompts marked Blueflame AI required involve searching, reading, comparing, summarizing, or drafting from data room document content, and require the project to have Blueflame AI enabled. All actions taken by AI, whether Blueflame AI or our MCP Server, follow existing Datasite permissions and authentication. Responses are grounded only in documents that the user has permission to access. AI output should be reviewed by advisors before relying on it.

 

• Sample prompts to get started
• Admin and setup
• MCP Specific
• Buy side review
• Blueflame AI prompts by topic
  • Financial & quality of earnings
  • Commercial and market
  • Legal, contracts, and IP
  • Technology and security
  • People and organization
  • Synthesis, risk, and investment committee
• Lifecycle skill alignment
• Combine prompts to accomplish larger tasks

Sample prompts to get started

Deal stage	Prompt	Best used by	Blueflame AI required
Pre-launch	Are there any contracts with unusual termination clauses, change-of-control provisions, or customer concentration risks?	Admin / Deal team	Yes
Pre-launch	What key financial, legal, HR, tax, or commercial documents appear to be missing from the data room?	Admin / Deal team	Yes
Pre-launch	What are the top 10 questions a buyer would likely ask based on the materials in this data room?	Admin / Deal team	Yes
Pre-launch	Summarize all leases in the data room, including expiry dates, renewal options, annual rent, and any change-of-control provisions.	Admin / Deal team	Yes
Active diligence	What EBITDA adjustments did the QoE team support versus what management originally claimed? Flag anything a buyer would challenge.	Admin / Deal team	Yes
Active diligence	Does the CIM show any metrics differently than the QoE report? Flag differences for the same period and explain why they may differ.	Admin / Deal team	Yes
Active diligence	Do any customer contracts automatically terminate if the company is sold? Flag the customer, revenue or ARR at risk, and what is needed to resolve it.	Admin / Deal team	Yes
Management prep	The buyer has noted discrepancies between the CIM and QoE report. What numbers look different, why do they differ, and what should the CFO be ready to explain?	Admin / Deal team	Yes
Management prep	Table every C-suite executive — base salary, bonus target, equity grant, and what happens to unvested equity if they are let go after close. Include any key person insurance.	Admin / Deal team	Yes
Management prep	How accurate has management been at hitting annual budgets historically? Based on the most recent quarterly actuals, are they on track to hit the current year plan?	Admin / Deal team	Yes

Admin and setup

Deal stage	Prompt	Best used by	Blueflame AI required
Material prep	Help me organize and categorize the source documents for this data room project.	Admin	No
Material prep	Suggest an optimal folder structure based on the documents we have collected and the type of deal we are preparing for.	Admin	No
Material prep	Review the files in this project and suggest clearer, more consistent document names.	Admin	No
Material prep	Identify documents that appear to be duplicates, outdated versions, or named inconsistently.	Admin	No
Project setup	Set up a new data room for this deal and recommend a folder index appropriate for the transaction.	Admin via MCP	No
Project setup	Create a numbered folder structure for a sell-side diligence process, including financial, tax, legal, HR, commercial, operational, and IT sections.	Admin via MCP	No
Project setup	Help me place the documents I have collected into the right folders in the data room.	Admin via MCP	No
Project setup	Invite these users to the project, assign the correct roles, and let me review the invitations before anything is sent.	Admin via MCP	No
Pre-launch	Which folders are empty, sparse, or appear incomplete before we open the data room?	Admin	No
Pre-launch	Review the data room and tell me whether it appears ready to launch. Highlight missing sections, quality issues, and major risks.	Admin / Deal team	Yes

MCP specific

Deal stage	Prompt	Best used by	Blueflame AI required
Portfolio view	Show me my active Datasite projects.	Any user via MCP	No
Project setup	Set up a new data room for Project Falcon. It is a mid-market business sale. Recommend the initial folder structure.	Admin via MCP	No
Project setup	Build a folder index for a sell-side diligence project and tailor it for a software company.	Admin via MCP	No
Project setup	Create the approved folder structure in the data room.	Admin via MCP	No
File management	Review file names in the data room and suggest standardized names that follow a consistent deal convention.	Admin via MCP	No
File management	Which sections of the data room are missing documents or appear underpopulated based on the folder index?	Admin via MCP	No
User management	Invite the following advisors as project admins and confirm when the invitations are ready to send.	Admin via MCP	No
Question management	Show me the open buyer questions, group them by category, and identify which ones need attention first.	Admin via MCP	No
Answer management	Draft responses to the open buyer questions using only the documents in the data room.	Admin via MCP	Yes
Content search	Are there any red flags in the financial statements, customer contracts, or employee agreements?	Admin / Reviewer via MCP	Yes

Buy side review

Deal stage	Prompt	Best used by	Blueflame AI required
Initial review	How much of this company’s revenue is recurring versus one-time or potentially at risk? Summarize based on available documents.	Buyer / Reviewer	Yes
Initial review	Summarize AR aging, AP aging, and deferred revenue for the most recent period. Flag anything unusual versus prior periods.	Buyer / Reviewer	Yes
IC prep	What EBITDA adjustments did the QoE team support, what did management claim, and what items were rejected? What is the most defensible number for a buyer to model?	Buyer / Reviewer	Yes
IC prep	Identify the company’s largest customers, revenue contribution, contract terms, renewal status, and any concentration risk.	Buyer / Reviewer	Yes
Legal diligence	For each material agreement, identify whether a change of control requires consent, triggers termination, or creates a payment obligation.	Buyer / Reviewer	Yes
HR diligence	For each executive, what equity acceleration is triggered on a change of control — single trigger or double trigger? What is the total potential cost to the buyer?	Buyer / Reviewer	Yes
Commercial diligence	What do the available documents show about customer churn, retention, renewals, and expansion? Flag any inconsistencies across materials.	Buyer / Reviewer	Yes
Financial diligence	Compare historical budget accuracy to the current forecast. Is management on track to hit the current year plan?	Buyer / Reviewer	Yes
Compliance diligence	Identify regulatory, licensing, privacy, or compliance obligations that apply to the target. Flag missing documentation or apparent non-compliance.	Buyer / Reviewer	Yes
Final bid prep	Summarize the top 10 risks a buyer should consider before submitting a final bid. Group them by financial, legal, commercial, operational, and HR risk.	Buyer / Reviewer	Yes

Blueflame AI prompts by topic

Paste a prompt into the Datasite Blueflame AI assistant and swap the named documents for your deal's file names, the assistant is fastest when you point it at a specific file. Every prompt already does the four things that matter: names the place, names the analysis, demands citations, and flags estimates. 

Financial & quality of earnings

Type	Use this for	Prompt
EBITDA Quality Bridge	Understand how reported EBITDA was adjusted and evaluate earnings quality.	In the QoE report, walk me from reported EBITDA to adjusted EBITDA for the last full year. List every add-back with its amount, cite the page, and flag any add-back you'd consider aggressive or non-recurring-in-name-only.
ARR / Revenue Bridge	Analyze the drivers of recurring revenue growth, expansion, contraction, and churn.	Build last year's ARR bridge—beginning ARR, new, expansion, contraction, churn, ending ARR—from the data room. Cite the source file and tab, and flag any component you had to derive rather than read directly.
Net Working Capital	Review working capital trends and evaluate the proposed working capital peg.	From the QoE or the databook, show net working capital by month for the last twelve months and the proposed target/peg. Cite the tab. Explain the drivers of any swing, and flag whether the peg looks buyer- or seller-favorable.
Revenue Durability	Determine how much revenue is recurring versus one-time and identify quality concerns.	What share of revenue is recurring versus one-time, by year? Cite the source file. Flag any revenue management counts as recurring that you'd treat as one-time, and say why.
Retention by Cohort	Evaluate customer retention trends and identify which segments drive performance.	Pull net and gross retention by cohort from the data room and describe the curve. Cite the file. If retention is stratified by product, segment, or payments attach, show the split and identify which segment drives the average.

Commercial and market

Type	Use this for	Prompt
Market Size	Validate the company's TAM/SAM and understand where the estimates come from.	What is the company's serviceable addressable market, and where does the figure come from? Cite the source, and tell me whether it's a management estimate or an independent third-party study.
Customer Concentration	Measure customer concentration risk and identify dependence on large accounts.	Show revenue (or ARR) concentration: the top 10 and top 20 customers as a percentage of the total. Cite the file. Flag any single-customer or single-cohort concentration risk.
Win/Loss Analysis	Understand competitive positioning and identify gaps in the diligence materials.	Summarize the documented win/loss reasons and the competitive set. Cite the file. Separate what's evidenced from what's anecdotal, and explicitly flag what the data room does not cover on competition.
Pricing Power	Evaluate historical pricing actions and their measurable impact.	Summarize pricing actions over the last two years and their revenue impact. Cite the source. Flag where the impact is measured versus estimated.

Legal, contracts, and IP

Type	Use this for	Prompt
Change-of-Control Review	Identify contracts requiring consent in a transaction.	Scan the top customer contracts for change-of-control or assignment-consent provisions. List each customer, the specific clause, and the file and page. Flag any reseller or customer agreements that are verbal or unsigned.
Material Contract Summary	Build an inventory of key commercial agreements and obligations.	List the most material third-party contracts (top customers, key vendors, licensors). For each: counterparty, term, renewal/termination, and any required consents. Cite each file.
IP & Licensing	Review intellectual property ownership and third-party dependencies.	Summarize the registered IP (trademarks, patents, key licenses). Cite the source. Flag any product capability that depends on a third-party license or an unsigned agreement.
Litigation Review	Summarize pending disputes and assess potential legal exposure.	Summarize any pending or threatened litigation and material disputes. Cite the file. Give the company's stated position and the potential exposure for each.

Technology and security

Type	Use this for	Prompt
Security Assessment	Review cybersecurity posture and identify the highest-priority risks.	Summarize the company's security posture—SOC 2 status, penetration tests, and the most material vulnerabilities—citing each source file. Then name the single biggest technology risk and explain why.
Architecture Review	Understand the technology stack and identify technical debt.	Summarize the product architecture across the company's products/modules—language, database, and hosting for each. Cite the tech DD or architecture files. Flag the least modern stack and the integration risk it creates.
Platform Modernization	Assess migration progress and execution risk.	What is the status of any platform migration or modernization effort? Cite the source. Give the pace, percent complete, and projected completion—and flag the risk if it slips.
Reliability & Scalability	Evaluate operational resilience and ability to support future growth.	Pull uptime, scalability, and any outage history from the technology diligence report. Cite the file. Flag reliability concerns that could affect the growth plan.

People and organization

Type	Use this for	Prompt
Headcount Analysis	Understand organizational structure and staffing by function.	Open the Employee Census and give total headcount by function and by department. If any figure is derived rather than stated in the file, say so.
Executive Employment Terms	Review executive compensation, retention, and change-of-control provisions.	Identify the key executives and their employment terms—base, bonus, equity, non-compete, and change-of-control/severance. Cite each agreement. Flag the biggest key-person risk.
Compensation Programs	Summarize employee compensation and incentive plans.	Summarize the compensation and benefits structure and any retention or incentive plans (option pool, MIU/PIU, 401k). Cite the file.

Synthesis, risk, and investment committee

Type	Use this for	Prompt
Risk Register	Create an investment committee-ready summary of the most significant diligence risks.	Across the CIM, the QoE, and the technology diligence report, give me the top 5 diligence risks as an IC-ready risk register: each row = the risk, the supporting evidence with its source file, and a severity rating. Flag where the sources disagree.
Cross-Document Reconciliation	Resolve conflicting numbers reported in different diligence documents.	Reconcile [the key metric] across the CIM and the QoE. Give both figures with their sources, and explain exactly what drives the difference.
Red Flag Review	Identify issues most likely to impact valuation or deal execution.	Sweep the data room for red flags: verbal or unsigned contracts, related-party items, aggressive add-backs, customer concentration, and pending litigation. Cite each finding and rank them by severity.
Management Questions	Generate follow-up diligence questions based on missing information.	Based on what is NOT in the data room, draft the top 15 follow-up questions for management, grouped by workstream. Be specific about which document or data point is missing.
Investment Summary	Create a first draft of an investment committee or deal summary.	Draft a one-page investment summary: thesis, key financials, top 3 strengths, top 3 risks—each point cited to a source file. Show me the outline before you draft the full page.

 

Lifecycle skill alignment

Deal lifecycle stage	Relevant AI capability
Material preparation	Smart file renaming, document organization, duplicate identification, folder planning
Project setup	VDR index setup, project creation, folder creation, user invitation workflows
Pre-launch readiness	Data room gap analysis, document quality check, launch readiness review
Active diligence	Risk analysis audit, content search, contract review, financial comparison
Buyer review	Revenue quality, working capital review, EBITDA analysis, contract and HR risk review
Management presentation prep	CFO call prep, discrepancy analysis, forecast credibility review
Signing and closing	Change-of-control review, consent tracking, equity acceleration analysis

Combine prompts to accomplish larger tasks

The prompts in this library are designed to work together. While each prompt can answer a specific question, you'll often achieve better results by using several prompts in sequence. This allows you to move from finding information, to analyzing it, to creating a polished deliverable.

Example: Responding to a buyer question

A buyer asks:

Can you summarize the company's major customer concentration risks
and identify any changes over the last three years?

Rather than trying to answer everything with a single prompt, break the task into a series of steps.

1. Find the relevant information
   Use a prompt to identify the documents that discuss customer concentration, major customers, and related risks.
2. Extract the key details
   Summarize the customer concentration risks described in those documents.
3. Analyze the findings
   Compare the disclosures across the last three years. Identify changes, trends, or inconsistencies.
4. Prepare the final deliverable
   Create a concise buyer response or executive summary based on the analysis, citing the supporting documents where appropriate.

Continue the conversation

Each response becomes the starting point for your next prompt. Rather than beginning again, ask follow-up questions to refine the analysis, compare additional documents, validate findings, or change the format of the output.

For example, you can ask the assistant to:

• Explain why a risk changed over time.
• Compare information across financial statements, contracts, and management presentations.
• Draft a management briefing, or executive summary.

Using prompts as building blocks helps you move efficiently from document discovery to analysis to a finished work product while keeping your conclusions grounded in the source documents.

   

Disclaimer

Some prompts require Datasite with Blueflame AI to be enabled because they involve AI powered search, document interrogation, comparison, or drafting grounded in data room content. Datasite MCP follows each user’s existing Datasite permissions and authentication. Users can only access projects, documents, and actions they are already permitted to access in Datasite. AI generated outputs should be reviewed by appropriate legal, financial, tax, or other advisors before being relied upon.

 

Regional considerations

Due to AI (Claude/OpenAI) provider restrictions, users in Hong Kong and China will not be able to access the document interrogation capability within Datasite with Blueflame AI. All other features remain unaffected. Users outside of Hong Kong and China are unaffected and will continue to have full access to all Datasite with Blueflame AI capabilities.