# Tokenomics And Generation Strategy Receipt Pattern

This note defines compact receipt metadata for AI usage cost, context size, generation strategy, deterministic rendering, full-source resend policy, lazy generation, and build-time versus runtime AI posture.

Core rule: tokenomics metadata is a planning signal, not a billing guarantee, pricing model, vendor comparison, model benchmark, runtime approval, or production approval.

The receipt pattern exists to make correctness-matters AI operable, sustainable, monitorable, regenerable, validatable, and payable without adding a dashboard or runtime telemetry collector.

## Receipt Fields

- `source_size_estimate`: Estimated source bytes and approximate source tokens used as planning context.
- `context_pack_size_estimate`: Estimated context-pack bytes and approximate context-pack tokens for the artifact entry.
- `rendered_view_count`: Count of existing rendered package paths represented by the artifact receipt.
- `deep_core_view_count`: Count of rendered views classified as deep-core views when that classification is known.
- `standard_view_count`: Count of rendered views classified as standard package views.
- `light_view_count`: Count of rendered views classified as light views when that classification is known.
- `classified_only_count`: Count of rendered views restricted to classified-only use.
- `model_call_strategy`: Normalized strategy for whether a model call is manual, assisted, deterministic after assistance, deterministic-only, runtime-required, or unknown.
- `full_source_resend_policy`: Normalized posture for whether full source material is resent to a model.
- `lazy_generation_policy`: Normalized posture for whether outputs are prebuilt, on demand, manual-request-only, not applicable, or unknown.
- `deterministic_rendering_policy`: Normalized posture for whether final package outputs are deterministic, partially deterministic, runtime model-generated, manual, or unknown.
- `build_time_ai_use`: AI-use posture for build, design, review, transformation, or generation, using the shared build-runtime posture values.
- `runtime_ai_use`: Whether a live runtime model call is required for the represented package artifact surface.
- `build_runtime_ai_posture`: Combined build-time and runtime AI posture.
- `regeneration_trigger`: Condition that should cause context-pack or portable-output regeneration.
- `cost_risk`: Qualitative cost-risk posture: low, medium, high, unknown, or not applicable.
- `review_required_before_scaling`: Boolean indicating whether review is required before scaling generation or runtime use.

## Normalized Values

- `build_runtime_ai_posture`: `none`, `build_time_only`, `runtime_only`, `build_time_and_runtime`, `unknown`, `not_applicable`
- `model_call_strategy`: `manual_only`, `single_pass_ai_assisted`, `ai_assisted_then_deterministic_render`, `deterministic_regeneration`, `runtime_model_call_required`, `unknown`
- `full_source_resend_policy`: `never`, `initial_generation_only`, `on_source_change`, `every_generation`, `unknown`, `not_applicable`
- `lazy_generation_policy`: `all_outputs_prebuilt`, `on_demand_allowed`, `manual_request_only`, `not_applicable`, `unknown`
- `deterministic_rendering_policy`: `deterministic`, `partially_deterministic`, `model_generated_runtime`, `manual`, `unknown`
- `cost_risk`: `low`, `medium`, `high`, `unknown`, `not_applicable`

## Operating Rules

- Build-time AI use does not imply runtime model dependency.
- Runtime AI use requires stronger observability, receipt, and control expectations.
- Deterministic rendering is preferred where stable artifacts can be generated without runtime model calls.
- Full-source resend policy records handling expectations. It does not authorize sending source material to a model.
- Lazy generation policy records whether outputs are prebuilt or can be generated later. It does not create a runtime generator.
- Cost risk is qualitative and must not be converted into live pricing, vendor comparison, or benchmark claims inside this package.
- Unknown build-time AI posture means the package has not certified source drafting or review history. It does not mean no AI was used.

## Current Context-Pack Posture

- The generated context-pack files and portable exports are deterministic static package outputs.
- The package records `runtime_ai_use`: `none` for the represented static artifact surfaces.
- The package records `build_time_ai_use`: `unknown` where source drafting, design, or review history is not certified.
- The package records `model_call_strategy`: `deterministic_regeneration` for context-pack regeneration from approved source files.
- The package records `lazy_generation_policy`: `all_outputs_prebuilt` for represented static outputs.
- The package does not add a cost calculator, chatbot, database, dashboard, agent runtime, telemetry collector, external retrieval layer, permission engine, promotion-contract runtime, or knowledge graph.