Compare
Eight estates. The same five indicators.
Each estate below is a synthetic dependency graph — nodes, typed dependencies, seven architecture domains. The same engine scores all eight, so the differences between them are differences in the architecture, not in the measurement.
Nothing here is fetched. There is no request, no account, and no corpus behind this page: the numbers are computed from graphs committed to this repository, so the same inputs give the same readout every time. That is a weaker claim than a live assessment and it is the accurate one.
Estate
Lens
M&A Due Diligence
Two estates mid-merger — duplicate masters, two landing zones, and an integration PMO holding the seams together.
- Nodes
- 25
- Dependencies
- 16
- Domains populated
- 7 of 7
- Longest chain
- 3 nodes
- Dependency waves
- 3
Overview — The engine unmodified — no domain is weighted above any other.
- Domain Coverage100.0How many of the seven architecture domains this estate populates. Under a persona, its focus domains count double.Spread across estates 28.6
- Complexity Balance92.7How evenly work is spread across complexity classes, as normalized entropy. An estate where everything is critical scores low.Spread across estates 14.4
- Governance Density100.0Share of nodes sitting in the focus set, against a target of ten percentage points per focus domain.Spread across estates 54.5
- Cross-Domain Integration56.3Share of dependencies that cross a domain boundary. A siloed estate scores low here however strong it looks elsewhere.Spread across estates 61.0
- Parallelizability88.0Longest dependent chain by duration, against total nodes, inverted — a shorter chain leaves more work parallelizable.Spread across estates 19.8
Overview is the engine unmodified. Every persona below is measured against this column.
All eight estates — Overview
Method
What a lens can change, and what it cannot.
A lens re-scopes; it never re-scores.
Each persona weights architecture domains. The five indicators are properties of the whole graph. Those are not the same axis, so multiplying the third indicator by the third domain weight would be arithmetic on unrelated quantities — it would produce a number, and the number would mean nothing.
So a lens resolves to a focus set of domains, and only the two indicators that are defined over domains are re-parameterized by it. The other three are untouched, and the overall figure stays a plain unweighted mean rather than a weighted composite.
Both re-parameterized indicators reduce to the baseline.
Weighted Domain Coverage counts every domain the estate populates, with focus domains counting double. With no focus set every weight is one, and it is the unmodified coverage figure.
Focus Concentration measures the share of nodes inside the focus set against a target of ten percentage points per focus domain. With a focus set of governance alone it is exactly the governance density the engine already computes — which is why Overview and the Auditor lens agree on it.
11 lenses resolve to 9 distinct focus sets.
A domain joins a focus set when the lens weights it at least half again above neutral. At that threshold some lenses land on the same set and therefore produce identical numbers. That is stated here rather than left for a reader to discover and mistake for a broken control.
- no focus set — Overview
- SECURITY — CISO, Post-Incident Readiness
- DATA — CDO
- PLATFORM — CTO
- GOVERNANCE + SECURITY — Board Member, Regulatory Audit Prep
- GOVERNANCE — Auditor
- AI — AI Officer
- DATA + GOVERNANCE + PLATFORM — M&A Due Diligence
- AI + DATA — AI Deployment Readiness
Two domains no lens can reach.
The lens catalog weights five of the seven architecture domains. APPLICATION and CAPABILITY have no corresponding weight, so no lens can focus them and no lens ever will until the catalog gains one.
They still count toward coverage — they are simply never the domains counted double.
Want the live path instead, against a real corpus rather than a published graph? Watch the agent work →