{"id":955,"date":"2025-11-26T15:15:13","date_gmt":"2025-11-26T15:15:13","guid":{"rendered":"https:\/\/anapoly.co.uk\/labs\/?p=955"},"modified":"2025-11-26T15:17:13","modified_gmt":"2025-11-26T15:17:13","slug":"lab-note-building-a-three-layer-instruction-set-for-anapolys-work-in-chatgpt-projects","status":"publish","type":"post","link":"https:\/\/anapoly.co.uk\/labs\/lab-note-building-a-three-layer-instruction-set-for-anapolys-work-in-chatgpt-projects\/","title":{"rendered":"Lab Note: building a three-layer instruction set for Anapoly’s work in ChatGPT Projects"},"content":{"rendered":"\n

Transparency label: AI-assisted<\/strong>
This post was developed collaboratively. Alec set the purpose and structure; ChatGPT drafted, critiqued, and refined the narrative under strict governance. Alec reviewed and accepted each stage.<\/em><\/p>\n\n\n\n

\n\n\n\n

The three-layer instruction set<\/a> took shape through a sequence of decisions, clarifications, and course\u2011corrections within a ChatGPT Project. What follows is a reflective account of how it emerged.\u00a0<\/p>\n\n\n\n

1. Recognising a structural gap<\/h2>\n\n\n\n

We began with scattered ingredients: a mission statement, the value proposition, and the Lab Framework. Each document covered something important, but none told the AI what world it was working in<\/em>. The Three\u2011Layer Model of Context<\/strong> made the gap obvious. We had talked for months about business context as a top layer, but there was no single, authoritative statement that the AI could rely on.<\/p>\n\n\n\n

The realisation was that, without a coherent top layer, the AI would continue to drift between voices, assumptions, and roles. The need for a stable business\u2011context layer became unavoidable.<\/p>\n\n\n\n

2. Using the extended mission document to surface the essentials<\/h2>\n\n\n\n

To understand what the top layer must contain, we drafted an extended mission document<\/strong>. Writing it forced us to specify Anapoly\u2019s identity, boundaries, ethos, and tone in operational rather than literary terms.<\/p>\n\n\n\n

Amongst other things, we clarified:<\/p>\n\n\n\n

    \n
  • that Anapoly is exploratory, not consultative;<\/li>\n\n\n\n
  • that we work only with synthetic data;<\/li>\n\n\n\n
  • that our tone is plain and grounded;<\/li>\n\n\n\n
  • that we are not selling AI or performing expert evaluation;<\/li>\n\n\n\n
  • that transparency is a defining value.<\/li>\n<\/ul>\n\n\n\n

    The exercise exposed the core elements the AI would need if it were to behave as a consistent Anapoly collaborator. Those insights quickly became the skeleton of the business\u2011context layer.<\/p>\n\n\n\n

    3. Asking the decisive question: What else does the AI need?<\/em><\/h2>\n\n\n\n

    The next turning point came when Alec asked: given the mission, the value proposition, and the Lab Framework, what else does the AI still lack? The answer was longer than expected. Beyond the mission and methods, the AI needed:<\/p>\n\n\n\n

      \n
    • explicit organisational identity;<\/li>\n\n\n\n
    • a clear audience model;<\/li>\n\n\n\n
    • non\u2011negotiable values;<\/li>\n\n\n\n
    • boundaries on what Anapoly does not do;<\/li>\n\n\n\n
    • tone and communication standards;<\/li>\n\n\n\n
    • risk posture;<\/li>\n\n\n\n
    • definitions of quality;<\/li>\n\n\n\n
    • strategic intent.<\/li>\n<\/ul>\n\n\n\n

      This list turned a loose idea into a concrete specification.<\/p>\n\n\n\n

      4. Consolidating into one canonical business\u2011context block<\/h2>\n\n\n\n

      At this point, we faced a structural choice: leave the business context scattered across multiple documents, or merge them into a single canonical block. Alec chose consolidation. That removed ambiguity and ensured that every project would begin with the same fixed identity, values, and constraints. Once the consolidated block was drafted, the top layer of the instruction set effectively snapped into place.<\/p>\n\n\n\n

      5. Rebuilding the behavioural governance layer from first principles<\/h2>\n\n\n\n

      Anapoly’s existing governance notes had grown organically and were no longer fully aligned with the clearer business context. Alec asked for a complete rewrite. We replaced fragmented instructions with a behavioural layer defining:<\/p>\n\n\n\n

        \n
      • tone (plain, dry, concise);<\/li>\n\n\n\n
      • stance (critical, truth\u2011first, no flattery);<\/li>\n\n\n\n
      • interaction rules (ask when unclear, challenge lazy assumptions);<\/li>\n\n\n\n
      • risk handling (flag operational, ethical, or data\u2011protection issues);<\/li>\n\n\n\n
      • constraints (no hype, no verbosity, no softening of justified critique).<\/li>\n<\/ul>\n\n\n\n

        The most important element was the decision to adopt Strong Governance<\/strong> as the default. The behaviour is now predictable, sceptical, and aligned with Anapoly\u2019s ethos.<\/p>\n\n\n\n

        6. Adding a deliberate escape clause<\/h2>\n\n\n\n

        Strong governance is effective but inflexible. To avoid it becoming a straitjacket, we added a controlled override mechanism: a mandatory keyword (OVERRIDE:<\/code>) followed by natural\u2011language instructions. The override lasts exactly one turn.<\/p>\n\n\n\n

        7. Sharpening the task\u2011prompt layer<\/h2>\n\n\n\n

        With the first two layers established, the task\u2011prompt layer became straightforward. It defines how immediate instructions are handled:<\/p>\n\n\n\n

          \n
        • follow the task prompt as written;<\/li>\n\n\n\n
        • interpret it inside the constraints of the business context and governance layer;<\/li>\n\n\n\n
        • ask for clarification when needed;<\/li>\n\n\n\n
        • use project files only when explicitly referenced.<\/li>\n<\/ul>\n\n\n\n

          This aligns directly with the Micro\u2011Enterprise Setup Blueprint<\/strong>, which treats task prompts as the active layer atop stable configuration.<\/p>\n\n\n\n

          8. Assembling the final three\u2011layer instruction set<\/h2>\n\n\n\n

          Once the components were complete, we assembled them in order:<\/p>\n\n\n\n

            \n
          1. Business Context<\/strong> \u2014 Anapoly\u2019s identity, values, tone, boundaries, risk posture, and strategic intent.<\/li>\n\n\n\n
          2. Behavioural Governance Layer<\/strong> \u2014 strict rules for tone, reasoning, interaction, critique, and risk.<\/li>\n\n\n\n
          3. Task\u2011Prompt Layer<\/strong> \u2014 guidance for interpreting immediate instructions.<\/li>\n<\/ol>\n\n\n\n

            We added a short explanatory note to clarify how the layers fit together and how overrides work.<\/p>\n\n\n\n

            The final result ensures the AI behaves like an informed, grounded collaborator who understands Anapoly\u2019s mission, values, and constraints.<\/p>\n\n\n\n

            \n\n\n\n

            The outcome<\/h2>\n\n\n\n

            This process created a durable operating profile for all future Anapoly projects. The instruction set now:<\/p>\n\n\n\n

              \n
            • anchors the model in Anapoly\u2019s identity,<\/li>\n\n\n\n
            • constrains drift through strict governance,<\/li>\n\n\n\n
            • ensures tasks are interpreted consistently,<\/li>\n\n\n\n
            • and provides a clean override path when needed.<\/li>\n<\/ul>\n\n\n\n

              We now have a dependable foundation to build on \u2014 and a clear method for adapting it when Anapoly evolves.<\/p>\n\n\n\n

              <\/p>\n","protected":false},"excerpt":{"rendered":"

              Transparency label: AI-assistedThis post was developed collaboratively. Alec set the purpose and structure; ChatGPT drafted, critiqued, and refined the narrative under strict governance. Alec reviewed and accepted each stage. The three-layer instruction set took shape through a sequence of decisions, clarifications, and course\u2011corrections within a ChatGPT Project. What follows is a reflective account of how […]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[47,477,253,492],"class_list":["post-955","post","type-post","status-publish","format-standard","hentry","category-lab-note","tag-chatgpt","tag-context-engineering","tag-contextual-scaffolding","tag-three-layer-context-model"],"_links":{"self":[{"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/posts\/955","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/comments?post=955"}],"version-history":[{"count":3,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/posts\/955\/revisions"}],"predecessor-version":[{"id":958,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/posts\/955\/revisions\/958"}],"wp:attachment":[{"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/media?parent=955"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/categories?post=955"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/anapoly.co.uk\/labs\/wp-json\/wp\/v2\/tags?post=955"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}