Apex AnalyticaAPEX ANALYTICA
// FRAMEWORK

A canonical scoring framework for systemic fragility.

Ω-Fragility (ΩF) is a 0–10 score for any node in a critical system, built from five canonical pillars under one fixed weighting.

FORMULAΩF = wI·I + wR·R + wJ·J + wC·C + wT·Twhere Σw = 1.0
ΩF PROFILE · SAMPLE NODE
246810IIRREPLACEABILITY9.0RRESTORATION LATENCY8.0JJURISDICTIONAL HAZARD4.0CCASCADE LOAD7.0TTAIL DEPTH6.0ΩF7.42
// FROM DATA TO PILLARS
4 ENGINES → 5 PILLARS

The pillars aren’t hand-entered. Manifold’s four reasoning engines turn raw graph data into the five scores the composite consumes — each engine feeds specific pillars.

SPIRTES
Structure Discovery
feeds
C
TARSKI
Formal Verification
feeds
J
PEARL
Counterfactual Engine
feeds
IR
PARETO
Cascade & Tail Simulation
feeds
CT

Full engine breakdown on the product page →

// THE FIVE PILLARS
I · R · J · C · T
Iw = 0.25
PILLAR · I
IRREPLACEABILITY

How impossible to substitute, technically or commercially.

Captures the true cost of substitution: technical compatibility, certification windows, contractual lock-in, and supplier concentration. A node with I=10 cannot be replaced on any meaningful timeline.

BACKED · PEARL
Rw = 0.20
PILLAR · R
RESTORATION LATENCY

Time to restore equivalent capacity after catastrophic failure.

Models the recovery clock: lead times, capital requirements, regulatory approvals, and skill availability. A node with R=10 has effectively no near-term recovery path — failure is permanent on operational horizons.

BACKED · PEARL
Jw = 0.15
PILLAR · J
JURISDICTIONAL HAZARD

Sanctions, conflict, export controls, and regulatory exposure.

Encodes geopolitical and regulatory exposure: sanctions risk, conflict zones, export controls, dual-use restrictions. A node with J=10 sits at the intersection of multiple hostile regimes or under a single decisive lever.

BACKED · TARSKI
Cw = 0.25
PILLAR · C
CASCADE LOAD

Downstream impact depth: GDP, sectors, nonlinear propagation.

Quantifies the systemic blast radius: how many downstream nodes lose function, how deep the cascade runs, what nonlinear amplifiers (panic, hoarding, deleveraging) kick in. A node with C=10 is load-bearing for the system itself.

BACKED · SPIRTES + PARETO
Tw = 0.15
PILLAR · T
TAIL DEPTH

Distributional depth beyond VaR — fat-tail severity.

Goes past Value-at-Risk to characterize the depth of the tail: how bad is the bad case, conditional on being bad. Heavy-tailed processes routinely produce realizations multiple sigmas beyond historical VaR.

BACKED · PARETO
// ONE CANONICAL WEIGHTING
Σw = 1.0

The five pillar weights are fixed and identical across every domain. Manifold applies the same composite everywhere, so a ΩF of 7.0 on a semiconductor supply chain is directly comparable to a 7.0 on pancreatic biology.

What changes per domain is the vocabulary, not the math: on a biomedical graph “Irreplaceability” reads as mechanism rarity and “Jurisdictional Hazard” as regulatory exposure. The numbers stay on one scale.

// WHAT THE SCORE DRIVES
ΩSF · CONTAGION · BUFFER

Per-node ΩF is the input, not the output. Manifold aggregates the node scores into system-level readouts — the live signals on the Ω monitor that tell you how close the whole system is to failure.

ΩSF
SYSTEM FRAGILITY
∝ Σσᵢ + (100 − buffer)

Live 0–100 system fragility index. Rises with aggregate shock severity and depleted buffer, and drives the regime band (STABLE → CRASH) in the Ω monitor.

Ω-Contagion
CASCADE REACH
|{j : Lⱼ(failᵢ) > τ}|

Count of downstream nodes whose loss exceeds threshold τ when node i fails — measured across Monte Carlo cascade runs.

Ω-Buffer
TIME TO FAILURE
365 · (buffer/100)

Days from the current buffer to systemic failure: 365 nominal, compressing toward 3 at breach. The operational window for response.

// SEE IT IN PRACTICE

Watch ΩF scoring on a live graph.

Trial accounts include the full pillar breakdown, the canonical weighting, and system-level aggregation.