For each schema, list its key assumptions and the observable signs that each one no longer holds

Why This Is a Rule

Schemas are built on environmental assumptions that can silently stop holding. A hiring schema assumes a buyer's market for talent. A product schema assumes competitor X's pricing model. A career schema assumes the industry values Y. When these assumptions stop holding, the schema continues operating — confidently producing outputs based on conditions that no longer exist.

Environmental change triggers detect assumption violations before the schema fails. For each key assumption, you specify an observable indicator that the assumption has broken: "If competitor X raises prices by >20%" or "If hiring time-to-fill exceeds 8 weeks consistently" or "If industry job postings shift from requiring Y to requiring Z."

The indicators must be observable — things you can check or that arrive through your monitoring channels — rather than abstract conditions. "If the market changes" is unmonitorable. "If monthly active users in our segment decline 10% QoQ for two quarters" is monitorable.

When This Fires

• When documenting a new schema and wanting to future-proof it against environmental change
• During assumption registers (Track assumptions in a five-field register: claim, impact if false, evidence, status, next test) when linking assumptions to monitoring
• When a schema's assumptions haven't been checked in months
• After an environmental shift that may have invalidated multiple schemas

Common Failure Mode

Listing assumptions without specifying observable indicators: "This schema assumes the market is growing." How would you know if it stopped growing? Without an observable indicator, you can't monitor the assumption — you'll only discover it broke after the schema fails. Every assumption needs a paired indicator.

The Protocol

For each schema: (1) List its 2-3 key environmental assumptions — the conditions it depends on. (2) For each assumption, specify an observable indicator of violation: "Assumption [X] would no longer hold if I observed [specific, monitorable event]." (3) Add these indicators to your monitoring: check them during scheduled reviews (Match review frequency to volatility: weekly for fast-changing, quarterly for stable, instant for surprises) or set up automated alerts where possible. (4) When an indicator fires → the assumption may have broken. Investigate immediately and review the schema.