┌──────────────┐
│ F1: Mental │ "What kind of thing is this?"
│ Models │
└──────┬───────┘
│ primes
▼
┌──────────────┐
│ F2: Principles│ "What cannot change?"
│ (Constraints)│
└──────┬───────┘
│ constrains
▼
┌──────────────┐ ┌──────────────┐
│ F4: Tradeoffs │◄───│ F3: Failure │ "What did we choose,
│ (Decisions) │───►│ Modes │ and what breaks?"
└──────┬───────┘ └──────────────┘
│ checked by ▲
▼ │ validates
┌──────────────┐ │
│ F5: Review │───────────────┘
│ Questions │ "Does the system match the model?"
└──────┬───────┘
│ maps to
▼
┌──────────────┐
│ F6: Archetypes│ "Which known pattern does this fit?"
└──────┬───────┘
│ drawn as
▼
┌──────────────┐
│ F7: Diagrams │ "Who needs to see what?"
└──────┬───────┘
│ built from
▼
┌──────────────┐
│ F8: Components│ "Shared vocabulary across teams"
│ (Vocabulary) │
└──────┬───────┘
│ bounded by
▼
┌──────────────┐
│ F9: Laws │ "What holds regardless of intent?"
│ (Empirical) │
└──────────────┘
FAILURE PATH: Skip F2 or F3 → decisions lack structural
honesty → FM11 (Observability Blindness) at the org level
The nine frameworks form a directed traversal, not a menu. Each framework narrows the space for the next: F1 determines which mental models are relevant, F2 identifies the constraints those models cannot escape, and F3 names the failure modes those constraints create exposure to. Skipping a framework does not save time — it defers the cost to the point where the skipped analysis would have changed the outcome, which is almost always during an incident at 2am.