Affective Governance in Adaptive Systems: An integrative framework for access to adequate change

An adaptive system can perform its present task while losing access to the changes its next task requires. The Affective Gain Module (AGM) programme studies six interfaces: internal-state readout and intervention, boundary-crossing time and intervention deadlines, context-dependent response, indicator integrity, shared-resource governance, and history-dependent loss of evidence and correction routes. This paper organizes them around access to adequate change: from available evidence, one admitted continuation reaches an adequate change for every compatible situation by the original deadline. Each companion supplies a coordinate, and E the rules for information sets, ledgers and effect records. Their joint contract matters because the interfaces share budgets and one interface's operations can change another's laws. A finite characterization inherited from E makes access exact within that contract; a composition statement gives sufficient conditions for local models to support it. A premise audit records derivation-relevant failures, their owners and candidate origins, retaining uncertainty when records are incomplete. Temporal priority is not causal identification. In an exact finite illustration, joint allocation preserves completions lost by rules strengthening one interface, and an injected link fault passes the declared companion-local checks but can be identified and repaired without extra supply. These are consequences of stipulated contracts, not measured affective effects; incremental value remains a prospective fault-injection hypothesis. Technical companion (included in this record). Independent Checking of Policy Certificates for AGM (Version 1.0, 1 October 2026) specifies a policy certificate for the Flagship's frozen finite example, FL-M1, that a separate program can check without trusting the planner that found the policy. The supplied archive contains 74 accepted policies for the original 96-cell grid (26 in the wide-scale family and 48 in the narrow-scale family); independent evaluation covers all 296 stipulated hidden histories and 1,532 primitive world operations. The remaining 22 cells have no positive certificate and receive no independent impossibility claim. These reproduce existing model outcomes through a new checking interface; they are not new empirical observations. Note on this release. Version 1.0 of the AGM Flagship (dated 24 September 2026; about 22,700 words) and of the technical companion (about 4,300 words). Files: the two manuscripts as PDF and one shared supplement archive (89 files) with the FL-M1 reference model and planner, the certificate data and the checking procedures; Python and its standard library are sufficient to rerun it. The results are consequences of stipulated finite contracts and completed computational checks. They are not measured affective effects, and completed computational checking is distinct from external scientific validation. Series. This is the flagship of the Affective Gain Module (AGM) programme. The six companion papers are archived separately: Paper A (affective state as control variable), Paper B (Quotient Critical Geometry), Paper C (terrain-contingent gain), Paper D (distortion-corrected emotional criticality), Paper E (affective endurance and closed-loop access to adaptive change) and Paper F (the Rigidity Trap). AI use disclosure. Generative AI (GPT-6.0, OpenAI; Claude Opus 5.5, Anthropic) was used substantively in preparing this work, including source comparison, drafting and editing, and, where applicable, mathematical and counterexample checks and the writing and running of supplementary code. The research questions, framework and final claims were directed and reviewed by the author, who takes full responsibility for the content, including the accuracy of all references and reported numbers. Repository metadata were prepared with assistance from Claude (Anthropic).