Data Room Agent

SkillCommerce & finance

Structures a virtual data room with a full folder taxonomy and staged access tiers, use it when you must organize diligence materials for buyers.

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the Data Room Agent skill

What this skill tells your AI

The instructions your AI receives, as published by andreworia/claude-finance-skills in packs/investment-banking/skills/data-room/SKILL.md and read by ahel’s review.

When to use

Use this when preparing a sell-side process and you must organize diligence materials so multiple buyers can review them in a controlled way. A disorganized data room slows diligence, invites repetitive questions, and leaks sensitive information too early. This agent gives you a clean taxonomy and a staged-access plan before documents go live.

What it does

It produces a data room index and an access-tiering plan: a complete folder structure for diligence materials plus a mapping of which documents are exposed in round one, round two, and confirmatory diligence, with watermarking and Q&A discipline built in.

Method

  1. Build the top-level folder taxonomy.

    • Standard folders: corporate, financial, commercial, legal, tax, HR, IT and technology, and material contracts.
    • Add industry-specific folders only where the sector demands them, for example regulatory or environmental.
  2. Populate each folder with standard subfolders.

    • Corporate: org chart, cap table, board minutes, constitutional documents. Financial: audited statements, management accounts, budget and forecast, working capital.
    • Commercial: customer and revenue analysis, pipeline. Legal: litigation, IP, permits. Tax: returns and structure. HR: org, comp, key contracts. IT: systems and security. Contracts: top customer and supplier agreements.
  3. Assign each document a sensitivity level.

    • Levels: open, sensitive, and clean-team only.
    • Clean-team items include customer-level pricing, named-customer margins, and detailed forward pipeline that a competitor could weaponize.
  4. Define staged access tiers.

    • Round one: overview and non-sensitive materials to all qualified bidders. Round two: detailed financials, key contracts (redacted where needed), and management-access materials to shortlisted parties. Confirmatory: remaining and clean-team materials to the preferred bidder.
    • Good looks like: exposure escalating strictly with buyer commitment, never ahead of it.
  5. Map every document to the earliest round in which it is released.

    • Honor its sensitivity level so a clean-team item cannot surface in round one even if it lives in an early folder.
  6. Set watermarking rules.

    • Apply dynamic watermarks carrying viewer identity and timestamp on all sensitive and clean-team documents.
    • Disable download or restrict to view-only where the item is highly sensitive.
  7. Establish Q&A discipline.

    • Route all questions through a single logged channel, tagged by workstream, with a tracked status and a standard turnaround.
    • Common pitfall: side-channel answers by phone that never get logged; prohibit them so every response is captured and consistent across bidders.
  8. Define permission and audit hygiene.

    • Use named user groups per tier, review the access log regularly, and set a downgrade or takedown path if a party drops out mid-process.

Inputs

  • The list of diligence documents available, or the categories you can supply
  • Which items are competitively sensitive or clean-team only
  • The number of process rounds and the gating between them
  • Buyer groups and how many parties are expected per round
  • Watermarking and download restrictions you require
  • Any redaction rules for third-party contracts
  • The Q&A turnaround target and the workstream owners

Output format

  • Data room index: the folder taxonomy with subfolders and the documents mapped into them, each tagged with its sensitivity level (open, sensitive, clean-team only).
  • Access-tiering plan: for each round (round one, round two, confirmatory), the buyer group, the folders and documents exposed, and the watermarking rules applied.
  • Q&A rules: channel, routing by workstream, turnaround, and the no-side-channel discipline.
  • Audit rules: user groups per tier and the access-log review cadence.
  • Describe all mappings as structured text, never a rendered markdown table.

Example

Meridian Foods sell-side. Index excerpt, Financial folder: audited statements (open), management accounts (sensitive), customer-level margin bridge (clean-team only). Access plan: round one exposes corporate overview, audited statements, and market materials to all six qualified bidders, view-only with dynamic watermarks. Round two adds management accounts and redacted top-20 customer contracts to the three shortlisted parties, with download disabled on the contracts. Confirmatory opens the customer-level margin bridge to the preferred bidder under a clean-team agreement, viewable only by named advisers. Q&A runs through one logged channel with a two-business-day turnaround, routed by workstream, and phone answers are prohibited so all bidders see the same record.

Signals

GitHub stars
20
Forks
4
Last commit
Sep 2026

Others that do the same job

Advanced
Item type
skill
Key
data-room-andreworia
Source
github.com/andreworia/claude-finance-skills