Computational Self-Regulation: A Field Definition and a Reference Implementation

Presley Jason Oakes — support@theevolved.net — September 10, 2025

Academic field: Computational Self‑Regulation (CSR) • Industry category: Self‑Mastery Operating Systems (SMOS) • Reference app: Capstone OS

Abstract

We define Computational Self‑Regulation (CSR) as the study and engineering of closed‑loop systems that measure, plan, act, and adapt across human ability domains to improve individual outcomes. We present a general architecture, a unified ontology with interpretable 1–100 anchors, a policy‑driven action layer, and a behavior‑shaping visualization (the “Mastery Map”). We argue that contemporary LLMs and modern cloud primitives make CSR feasible at consumer scale. We outline initial research tasks and propose an open Aptitude Interchange Format (AIF) to standardize data exchange.

1. Introduction

Most consumer “self‑improvement” tools stop at data collection, generic advice, or habit tracking. CSR integrates these components into a single feedback loop: assess → summarize → plan → act → log → adapt → visualize. The reference implementation, Capstone OS, demonstrates the loop across seven domains: Cognition, Fitness, Movement, Health, Senses, Social, and Emotions.

2. Related Work

• Psychometrics (Spearman, Cattell, Big Five) — rigorous measurement, limited cross‑domain synthesis.
• CBT & behavioral medicine — structured interventions with clinical scope.
• Deliberate practice & periodization — micro‑cycles to move capabilities.
• Quantified Self & wearables — abundant signals, weak plan integration.
• HCI & behavior change — interfaces that scaffold action.
• Modern LLMs — plan synthesis from structured profiles and context.

3. CSR Architecture

The CSR loop composes a measurement layer (interpretable scores), a reasoning layer (AI Advisor + policy engine), an action layer (quests/goals/tools), and a visualization layer (Mastery Map). Provenance records models and policy versions for auditability.

4. Ontology & Scales

Each domain has subscales with descriptive anchors for 0–100 (in 5‑point steps). Anchors enable self‑assessment and make AI plans interpretable. Example (Cognition): Attention, Memory, Language, Perception, Problem‑Solving, Decision‑Making, Creativity.

5. Planning: AI Advisor + Policy

The Advisor merges profile, assessments, history, and context to propose goals and micro‑cycles. A simple policy prioritizes weakest‑domain work tempered by recovery state (e.g., sleep, stress). This yields 2–3 high‑leverage quests per day.

6. Action: Quests, Goals, Tools

Quests are concrete behaviors (effort‑bounded). Goals aggregate completions over rolling windows (e.g., 7 days). Tools (drills, workouts, finance tasks) operationalize the plan.

7. Visualization: Mastery Map

A radar/terrain hybrid maps capabilities into a spatial metaphor (“the mountain”). Progress is visible and motivating; users can “see” themselves climbing.

8. Ethics, Privacy, & Safety

• Consent for data use and research.
• Privacy by default; export/delete controls.
• Explainable plans; human‑in‑the‑loop for sensitive domains.
• Guardrails and safe failure modes.

9. Research Program

• Reliability: internal consistency and test–retest.
• Validity: convergent validity with existing scales.
• Intervention pilots: 7–14 day micro‑cycles; effect sizes and adherence.
• Visualization efficacy: does the Mastery Map change behavior?

10. Aptitude Interchange Format (AIF) Summary

AIF is a JSON schema for CSR data: profiles, assessments, context, plans, quests, completions, outcomes, and provenance. See spec/aif-schema-v0.1.json.

11. Reference Implementation

Capstone OS integrates the CSR loop end‑to‑end. Key innovations: unified anchors, policy‑driven quests, and a motivating spatial visualization.

12. Conclusion

CSR formalizes closed‑loop personal improvement. With a common ontology and spec, researchers and builders can collaborate on benchmarks, datasets, and safer, more effective systems.

Cite as: Oakes, P. J. (2025). Computational Self‑Regulation: A Field Definition and a Reference Implementation. DOI: https://doi.org/10.5281/zenodo.17095406.

Links: CSR site • TheEvolvED.net • Capstone OS