Red Flag Register

SkillCommerce & finance

Produces the single register in which every diligence finding is graded by severity and classified as priceable, fixable or walk-away, with a named owner and a resolution status, for when you need each flag to land in the price, the SPA or the value creation plan before IC.

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 Red Flag Register skill

What this skill tells your AI

The instructions your AI receives, as published by andreworia/claude-finance-skills in packs/private-equity/skills/red-flag-register/SKILL.md and read by ahel’s review.

When to use

Use this skill from the first advisor read-out through to signing. The issues log tracks that a finding exists; the red flag register decides what is done about it. Every finding takes one of three routes -- it is paid for, it is fixed, or it ends the deal -- and forcing that choice early is what stops forty amber items reaching IC with no aggregate stated and no owner against any of them.

What it does

Maintains one register in which each finding carries a quantified impact, a disposition of priceable, fixable or walk-away, the mechanism that carries it, a named owner, and a status that closes only when the finding has landed somewhere real.

Method

Step 1 -- Keep one register, not one per advisor

Findings arrive from QoE, commercial, legal, tax and IT in five formats. Restate each in the fund's own wording and number it once. Parallel advisor lists are how a finding gets discussed three times and resolved never.

Step 2 -- Grade by quantum, not by adjective

Every entry carries a number: recurring EBITDA effect, one-off cash cost, or contingent exposure with a probability against it. Convert recurring effects into turns at the entry multiple so severity is comparable across workstreams. "Significant" is not a grade.

Step 3 -- Classify the disposition

Priceable: the exposure is bounded and quantifiable, so it can be taken in the bid, an escrow, a holdback or a completion adjustment. Fixable: it is remediable at known cost within a known period, so it becomes a condition precedent, a warranty, a W&I placement or a 100-day workstream. Walk-away: the downside is not bounded, or the fix depends on a third party outside the deal's control. The test for walk-away is boundedness, not size -- a large bounded number is priceable, a small unbounded one is not.

Step 4 -- Route each flag to its mechanism and its owner

Name the instrument that will carry the finding and the individual accountable for putting it there -- not a firm. Legal owns the CP wording, the deal VP owns the bid adjustment, the operating partner owns anything becoming a 100-day item. A finding with a disposition but no mechanism is still open.

Step 5 -- Run the aggregate and correlation test

Sum the priceable column and set it against bid headroom. Then test correlation: findings driven by one cause -- a single regulator, a single ERP migration, a single customer -- are one exposure, not four, and cannot be assumed to fail independently.

Step 6 -- Close only on landing

Status runs open, quantified, routed, landed. Landed means the number is in the bid letter, the clause is in the SPA draft, or the action is in the 100-day plan with a budget. A flag marked resolved because it was discussed is not resolved.

Step 7 -- Hand the register to IC and to the SPA

The IC risk matrix is the register filtered to material entries; the CP schedule and the indemnity list are the register filtered to the fixable ones. Anything walk-away goes to the deal partner the day it is graded, not at papers.

Inputs

  • Advisor read-outs and red-flag reports, all workstreams
  • QoE EBITDA, entry multiple, bid headroom to the return threshold
  • Information gap log, so unevidenced items are not graded as findings
  • SPA draft status and W&I broker position
  • Named deal team and advisor owners

Output format

  • Header: entry count by disposition, aggregate priceable amount, headroom
  • One entry per flag: description, workstream, quantum in EBITDA and turns, disposition, mechanism, owner, status
  • Correlation note: which flags share a single cause
  • Walk-away section, listed separately and dated
  • Filtered extracts for the IC risk matrix and the CP schedule

Example

Fictional target: Pellworth Care Services, 34 sites. QoE EBITDA GBP 18.0m, proposed entry 9.5x, opening bid GBP 171.0m.

RF-07, priceable. Agency staffing runs GBP 1.1m above the level supported by the accepted add-back, and the rate shows no sign of normalising. Recurring, so it is taken in the multiple: EBITDA to GBP 16.9m, bid to GBP 160.6m -- GBP 10.5m of price, 0.58 turns. Owner: deal VP. Status: landed in the revised bid letter.

RF-11, fixable. Two of the 34 sites are operating on lapsed registrations. Remediation is GBP 0.4m of one-off cost over about eleven weeks, taken as a completion-accounts adjustment rather than in the multiple, bringing the bid to GBP 160.2m, with re-registration of both sites as a condition precedent. Owner: legal counsel. Status: routed.

RF-02, walk-away test. An enforcement notice capable of suspending admissions across the estate is not priceable, because the revenue at risk is the estate rather than a site. It was escalated to the deal partner the day it was graded. The register's value is that it forces this distinction before the bid is revised, not after.

Signals

GitHub stars
20
Forks
4
Last commit
Sep 2026
Advanced
Item type
skill
Key
red-flag-register
Source
github.com/andreworia/claude-finance-skills