skills/buyer-investor-list/references/qa-checklist.md

# MD Review QA Checklist

Run this checklist before finalizing output.

## Completeness

- Are all relevant archetypes covered for the mandate?
- Are obvious strategics, sponsors, platforms, lenders, or investors included or explicitly excluded?
- Are existing lenders/shareholders/prior process participants considered when relevant?
- Are sponsor-backed platforms mapped, not just sponsor names?
- Are cross-border, family office, permanent capital, infrastructure, credit, or distressed buyers considered only where relevant?

## Relevance

- Does every tier 1 party have specific rationale?
- Are any names included only because they are large, famous, or broadly tagged to the sector?
- Does each party fit the transaction size, geography, sector, stage, and structure?
- Are stale or weakly supported parties labeled as watchlist/validate?

## Ability to transact

- Can each tier 1 party plausibly pay, finance, invest, or hold the required amount?
- Are fund size, vintage, mandate, check size, and control/minority fit considered for sponsors?
- Are hold size, lead capability, sector appetite, and documentation speed considered for lenders?
- Is aum avoided as a proxy for dry powder or check size unless supported?

## Process risk

- Are confidentiality-sensitive competitors flagged?
- Are antitrust/regulatory flags included where obvious?
- Are portfolio conflicts and commercial relationship sensitivities checked?
- Are leak/retrade/bad-process reputation issues reflected if known?
- Are client do-not-contact instructions respected?

## Outreach strategy

- Are waves logical given valuation, certainty, and confidentiality objectives?
- Is the teaser/nda/named-disclosure approach clear?
- Are client approval gates explicit?
- Is there a relationship owner or contact validation ask for priority parties?
- Does the output identify who an MD should call personally?

## Data quality

- Are parent/subsidiary/sponsor/platform duplicates resolved?
- Are prior notes, nda status, iois/lois, and outreach history preserved?
- Are sources and confidence levels included?
- Are facts, assumptions, and inferences separated?
- Are missing fields identified without hallucinating?

## Final "so what"

- Does the output clearly recommend who to contact first and why?
- Does it identify who not to contact and why?
- Does it explain how the list creates competitive tension or improves certainty?
- Could an MD defend this list in front of a client or committee?

## Standalone HTML Presentation

- Does an HTML-selected output follow `../../../references/html-artifact-standard.md` as a polished standalone buyer-universe report?
- Does the first viewport make the process posture, conditional first wave, high-value holds, and required approvals clear?
- Is there one central ranked universe table rather than several overlapping retellings of the same decision?
- Are held or excluded sensitive parties handled in the dedicated hold register rather than duplicated as full ranked-table rows, unless comparison is essential?
- Are pre-outreach validation and diligence gates consolidated into one actionable table?
- Are numerical scores demoted unless they materially explain prioritization or were requested?
- Are citations readable at the claim or row level rather than repeated across every cell?
- Are generic dashboard navigation, reader-action bars, related-file panels, internal contracts, and other visible generation machinery absent unless specifically useful and requested?
- Has the local HTML been rendered and visually inspected with local headless-browser screenshots before delivery?