CAL: A DSL Where the Cascade Is the Program

The Problem With Frameworks

Business analysis frameworks are everywhere. SWOT gives you four boxes — Strengths, Weaknesses, Opportunities, Threats. Porter's Five Forces gives you five competitive pressures. PESTEL gives you six environmental categories. You fill them in, you discuss them in a meeting, you move on. The analysis lives in a slide deck.

The problem is not the frameworks — it is what they cannot do. They describe dimensions in isolation. A quality issue goes in one box. A revenue concern goes in another. But organizations do not fail in isolation. A quality issue cascades into customer dissatisfaction, which cascades into revenue pressure, which triggers regulatory scrutiny, which strains operations, which drives staff turnover. The cascade is the event — and no framework traces it.

CAL does.

The Six Dimensions

CAL models organizations across six dimensions — not as independent categories but as a connected system. When one dimension moves, the others respond.

The typical failure cascade runs: Quality issue (D5) surfaces, customer trust erodes (D1), revenue contracts (D3), regulators take notice (D4), operations are restructured under pressure (D6), talent exits (D2). CAL traces that chain explicitly — not as an observation after the fact, but as a declared program that scores each step.

What a DSL Changes

The key word is syntax. CAL is not a scoring rubric you fill in — it is a language with a parser, a runtime, and deterministic evaluation. You declare the conditions, the dimensions, the thresholds. The runtime executes it. The same program produces the same result every time.

That distinction matters more than it sounds. When analysis lives in a framework, it is only as good as the analyst applying it on a given day. When analysis is a program, it is auditable, reproducible, and independent of who is in the room.

CAL is a five-layer pipeline, each layer encoded as keywords:

A Minimal CAL Program

Here is a complete CAL program — sensing for cascade conditions, measuring the gap, and producing a decision:

-- SENSE: find entities where conditions are elevated
FORAGE entities
WHERE sound > 7
ACROSS D1, D3, D5
DEPTH 2
SURFACE cascade_map

-- MEASURE: score the gap between methodology and performance
DRIFT cascade_map
METHODOLOGY 85
PERFORMANCE 35

-- DECIDE: produce an action score against a threshold
FETCH cascade_map
THRESHOLD 1000
ON EXECUTE CHIRP critical "Cascade conditions met — review D1, D3, D5"

FORAGE searches. DRIFT measures the gap — here, Methodology 85 minus Performance 35 equals a gap of 50, a significant teaching signal. FETCH produces a numeric score — Chirp × |DRIFT| × Confidence — and compares it against the declared threshold. If the score exceeds 1,000, the decision is EXECUTE. Below 500, it is WAIT.

No discussion required. The program produces a decision.

Production Evidence

CAL is not a proposed framework. It has been applied to 228+ case studies across 148+ sectors — banking, technology, geopolitics, healthcare, real estate, energy, sports, agriculture. FETCH scores range from 898 to 4,461.

The highest-scoring case is UC-039 — Silicon Valley Bank. Six of six dimensions compromised in 48 hours. FETCH score: 4,461. The cascade was traceable before the bank run accelerated — asset-liability mismatch, uninsured deposit concentration, an 18-month CRO vacancy. CAL maps the chain. The score reflects the severity.

The runtime is open source, published on npm, and carries a Zenodo DOI. The methodology is reproducible by anyone.

What Comes Next

This post is the introduction. Three deeper concepts are worth their own treatments:

Resources