Capacity-Safe Systems Standard

Every system built for humans was designed at full capacity. None of them work when the human operating them doesn't. The failure was never human. It was always architectural.

A Promise Before the Standard

If you have ever abandoned a planner, quit an app, failed to show up, disappeared from a system, or called yourself broken for not being able to keep up — this document is for you. You were not failing. You were in a capacity state that the system was not designed for. The failure was the design.

I am regulated, therefore I can.

A Warning Before the Standard Begins

This framework can be misused. Gentle colors and warm language do not make a distress-extraction system Capacity-Safe. The mark certifies structure, not aesthetics. It certifies demand architecture, not tone.

If distress drives your conversions, fear drives retention, or shame drives re-engagement — the Capacity-Safe mark is not for you.

Preamble

This standard exists because of a design failure so universal it became invisible. For generations, every system humans interact with was architected around one assumption: the human using it is available, regulated, resourced, and at capacity.

That assumption has never been true. The Capacity-Safe Systems Standard does not ask humans to adapt to broken systems. It asks systems to be designed around the actual range of human capacity — including its lowest points and its highest ones.

This is not accommodation. It is not therapy. It is not wellness culture. It is infrastructure.

Capacity cannot be demanded. It can only be restored.

Part I · Foundations

Capacity is the available bandwidth of the human nervous system — not a character trait, not willpower. Version 3.0 introduces the Capacity Alignment Principle: the system works for the human it is serving right now. Valid goals include return, stabilization, expansion, and sovereignty — not only “return to baseline.”

The standard applies to products, AI, workplaces, schools, hospitals, communities, and policy. It is an operational systems standard — not a clinical instrument.

Part II · The Capacity Map v3.0

Eight operational states (plus the Disconnected Function overlay). Fluid, bidirectional. No state is a failure.

Disconnected Function is an overlay — high performance without integration. The system must not exploit it.

Part III · The Infrastructure Stack

Four layers in order: Somatic → Cognitive → Relational → Operational. Most systems address only one layer incorrectly.

Part IV · The Recovery Spine

Session-level protocol: Safety → Offload → Simplification → Gentle Movement → Recovery Protection. The Spine may stop at any phase. Safety alone is enough.

Part V · The Recovery Arc

Longitudinal capacity path over days, weeks, and years — distinct from the session-level Recovery Spine.

Part VI · The Eighteen Capacity-Safe Principles

Version 3.0 expands from twelve to eighteen principles. All eighteen must be met for Foundation certification under v3.0. Read complete definitions in the full PDF.

1. Safety Before Function — No system may demand task completion before establishing perceived safety.
2. Capacity Is the Starting Point — Every design decision begins with what a human at low capacity can actually do.
3. Shame Is a Design Failure — Any feature or language that produces shame in a low-capacity user is a design error.
4. One Thing at a Time — Below Grounded state, the system presents one choice, one step, one action.
5. Stopping Is a Feature — Stopping must be a first-class user action with its own pathway and return protocol.
6. Language Scales With Capacity — Language simplifies as capacity drops. Short sentences, plain words, no abstraction at low states.
7. The System Does Not Judge — No system language may communicate judgment of the user's state, pace, absence, or choices.
8. Children and Adolescents Are Not Small Adults — Child-facing systems must be designed around their specific nervous system stage.
9. Recovery Is Not Passive — Recovery is an active state that requires specific design support. The Recovery Arc must be designed, not assumed.
10. Access Is Infrastructure — Capacity-Safe design is the base standard, not an accessibility add-on for a subset of users.
11. AI Must Know What State It Is Talking To — AI must detect or accept declared capacity state and modify responses accordingly.
12. Capacity Alignment Is the Goal — The system aligns with the user's actual capacity: return, stabilization, expansion, or sovereignty.
13. The Body Is the First Interface — Visual design, motion, sound, and spatial arrangement are somatic inputs before they are cognitive ones.
14. Trust Is Built Through Consistency — A system that behaves unpredictably trains the nervous system for vigilance, not rest.
15. The System Must Be Capable of Ending — Every Capacity-Safe system must have a designed end. Infinite systems are extractive.
16. Human Connection Is Not Optional — Every system must provide a visible, friction-free pathway to human connection.
17. The Unspeakable Does Not Require Disclosure — Systems must provide capacity support without requiring users to explain their internal state.
18. Temporal Justice — Deadlines and pacing must be adjustable proportional to detected or declared capacity.

Part VII · AI Systems Requirements

AI systems interact with humans across all capacity states, often at moments of acute distress. The scale of AI interaction makes this the highest-priority domain for Capacity-Safe compliance. AI carries special risk because it can simulate intimacy, produce infinite content, and interact during vulnerable moments at scale.

• AI must accept declared capacity state as a primary input shaping every subsequent response
• AI must never push a distressed user toward productivity before safety is established
• AI interactions must have a clearly designed end — infinite loops are not Capacity-Safe
• AI must be capable of directing to human support without friction or shame
• AI must not simulate attachment beyond what it can genuinely sustain
• AI must create after-work maps — surfacing hidden cleanup, checking, and emotional labor created by automation

Complete implementation requirements for digital products, workplaces, schools, healthcare, and communities are in the full PDF.

Part XIV · Capacity-Safe Certification

• Core — no prohibited patterns, state declaration available, one low-capacity mode with collapsed choices, visible exit, warm return, shame-free language audit passed. Entry point for small teams and early-stage products.
• Foundation — eighteen principles, Recovery Spine, basic Capacity Map response across all eight states
• Full Standard — complete Infrastructure Stack, Recovery Arc, Capacity Privacy, AI guardrails if applicable
• Advanced — measurable capacity alignment outcomes, consent-based detection, verified impact data
• Institutional — organization-wide implementation

Certification requires evidence, practitioner review, user-reported alignment data, and public commitment to the Capacity Alignment Principle — not the old “capacity return only” framing.

Certification paths · AI Agents practitioner credential

Closing · Why This Standard Exists

I am regulated, therefore I can.

Every person who abandoned a planner, quit an app, or called themselves broken was not failing. They were in a capacity state the system was not designed for. The failure was the design.

Human capacity is not a bug in the system. It is the system. Everything built for humans must be built around it.