Open-Core Decision Sheet

Purpose: Final internal decision layer. Forces concrete choices: what we commit open, commit paid, defer, or leave for later discussion. Not an audit or exploratory matrix—this is the decision artifact.

Open-Core Decision Sheet

Purpose: Final internal decision layer. Forces concrete choices: what we commit open, commit paid, defer, or leave for later discussion. Not an audit or exploratory matrix—this is the decision artifact.

Source: Distilled from docs/MONETIZATION_OPEN_CORE_AUDIT.md and docs/PRODUCT_MATRIX_OPEN_CORE.md.

Positioning:

AINL is the open language for agent workflows. We sell the operational layer: governance, visibility, managed runtime, deployment, and support.

Strategic rule:

Language legitimacy stays open.
Operator labor / governance / convenience gets monetized.

Paid buckets: Managed | Enterprise | Expertise. We do not position thin, easily cloned utilities as premium products.

A. Decision summary

| Outcome | What it means | |---------|----------------| | Commit open | We commit now: this stays open. Language legitimacy, core tooling, baseline docs/examples, onboarding/preflight, adapter discovery, basic policy validator, health envelope schema, and all simple scripts/tools. | | Commit paid | We commit now: these are commercial pillars. Managed (hosted runtime, managed alignment, managed dashboard/ops); Enterprise (governance/audit, deployment kits, policy packs, supported monitor pack, premium connectors with depth); Expert Services (support, onboarding, certification, consulting). | | Defer | We explicitly do not decide yet. Premium capability catalog, premium curriculum/certified packs, and any offering that is too early or too easy to clone before we have product/support maturity. | | Discuss later | Depends on future demand or execution reality. Revisit when we have signal (e.g. enterprise tool API, industry-specific curriculum). |

B. Decision table

| Surface / feature / area | Repo grounding | Decision | Paid bucket | Why | Replication risk | Notes / conditions | |--------------------------|-----------------|----------|-------------|-----|------------------|--------------------| | AINL spec, grammar | docs/AINL_SPEC.md, docs/language/grammar.md, docs/language/AINL_CORE_AND_MODULES.md, docs/reference/AINL_V0_9_PROFILE.md | Commit open | — | Language legitimacy | — | Non-negotiable | | Parser, compiler, canonical IR | compiler_v2.py, compiler_grammar.py, grammar_constraint.py, grammar_priors.py | Commit open | — | Single auditable compile path | — | Core of compile-once-run-many | | Reference runtime | runtime/engine.py, runtime/compat.py, runtime/values.py, SEMANTICS.md, docs/RUNTIME_COMPILER_CONTRACT.md | Commit open | — | Conformance and portability | — | Reference semantics | | Core runtime adapters | runtime/adapters/ (http, sqlite, fs, tools, wasm, base, replay) | Commit open | — | Baseline usefulness | — | Required for local use | | IR schema, graph schema | docs/reference/IR_SCHEMA.md, docs/reference/GRAPH_SCHEMA.md | Commit open | — | Public contract for ecosystem | — | Never paywall | | CLI (validate, run, basic emit) | cli/main.py, validate/run/emit entrypoints | Commit open | — | Local dev and CI | — | Essential workflow | | Baseline validation | scripts/validate_ainl.py, conformance tests, docs/CONFORMANCE.md | Commit open | — | “Does this compile?” essential | — | Never paywall validation | | Core docs | docs/DOCS_INDEX.md, ARCHITECTURE_OVERVIEW, INSTALL, CONFORMANCE, README.md | Commit open | — | Onboarding and trust | — | Transparency | | Baseline examples, golden | examples/golden/*.ainl, docs/EXAMPLE_SUPPORT_MATRIX.md | Commit open | — | Learning and training | — | Canonical path | | Adapter registry, manifest (core) | docs/reference/ADAPTER_REGISTRY.md, tooling/adapter_manifest.json, core in ADAPTER_REGISTRY.json | Commit open | — | “What can I call?” is language surface | — | Discovery stays open | | Capability registry (schema + canonical) | tooling/capabilities.json, capabilities.schema.json, docs/reference/CAPABILITY_REGISTRY.md (core subset) | Commit open | — | Tool API v2 and agent discovery | — | Open subset for open core | | Baseline eval harness, report schemas | Concepts and open report schemas | Commit open | — | Reproducibility and contract | — | Research and community | | Bot onboarding, preflight | docs/BOT_ONBOARDING.md, docs/OPENCLAW_IMPLEMENTATION_PREFLIGHT.md, tooling/bot_bootstrap.json | Commit open | — | Universal implementation discipline | — | Must stay free | | Basic policy validator | tooling/policy_validator.py | Commit open | — | Lightweight governance; safe default | — | Simple utility; keep open | | Health envelope schema | docs/operations/STANDARDIZED_HEALTH_ENVELOPE.md | Commit open | — | Interop for monitors/dashboards | — | Schema stays public | | Extension adapters (OpenClaw, agent) | adapters/openclaw_integration.py, adapters/agent.py, examples/openclaw/ | Commit open | — | Reference implementation; adoption | — | Paid = supported tier only | | Graph / IR tooling | tooling/graph_*.py, ir_canonical.py, ir_compact.py, step_focus.py, trace_focus.py | Commit open | — | Introspection; simple utility | — | Dev tooling | | Tool API v2 (base projection) | scripts/gen_tool_api_v2_tools.py, tooling/tool_api_v2.tools.json (open subset) | Commit open | — | Agent discovery | — | Base open | | Capability report/filter scripts | scripts/capabilities_report.py, scripts/capabilities_filter.py | Commit open | — | Discovery; simple utility | High | Paid = packaged + SLA, not script | | Memory contract | docs/adapters/MEMORY_CONTRACT.md | Commit open | — | Extension contract | — | Interop | | Memory tooling (retention, validate, import/export) | scripts/memory_retention_report.py, validate/import/export, tooling/memory_bridge.py, etc. | Commit open | — | Operator hygiene; simple scripts | High | Paid = governance suite packaging | | Operator-only audit script | scripts/operator_only_adapter_audit.py | Commit open | — | Visibility; simple script | High | Paid = audit suite packaging | | Coordination validator, mailbox validation | tooling/coordination_validator.py, scripts/validate_coordination_mailbox.py | Commit open | — | Envelope validation; reference | High | Paid = compliance packaging | | Autonomous ops examples, run scripts | examples/autonomous_ops/*.lang, demo/*.lang, scripts/run_*.py, docs/operations/AUTONOMOUS_OPS_MONITORS.md | Commit open | — | Reference monitors and patterns | — | Paid = supported pack / Managed | | Hosted AINL runtime / managed ops | Hosted service (no repo code) | Commit paid | Managed | Run at scale; uptime; operator labor | Low | Operational guarantee | | Managed alignment pipeline | Builds on run_alignment_cycle.sh, sweep, eval, run health | Commit paid | Managed | We run sweep, eval gate, trends; SLA | Low | Operational guarantee | | Managed dashboard / ops platform | Builds on health envelope, run health; aggregation, alerts | Commit paid | Managed | Visibility; regression; run health | Medium | Only if it provides aggregation, trends, regression visibility, alerts, governance integration, or SLA-backed operational value—not just UI over open outputs | | Enterprise governance / audit suite | Package: audit, capability governance, coordination validation, memory hygiene, policy packs | Commit paid | Enterprise | “Who can do what”; compliance | Low | Policy packs + integration | | Deployment kits (SSO, K8s, terraform, hardening) | Conceptual playbooks | Commit paid | Enterprise | Deploy in your VPC; reduce risk | Low | Reduces labor and risk | | Policy packs (curated, industry/compliance) | Builds on open policy validator | Commit paid | Enterprise | Compliance; “prove we’re in control” | Low | Curated + support | | Supported ops monitor pack | Curated autonomous ops + runbooks + health envelope + updates + SLA | Commit paid | Enterprise / Managed | Guaranteed updates and support | Medium | Examples stay open; support = paid | | Premium connector packs | Enterprise adapters beyond http/sqlite/fs/tools | Commit paid | Enterprise | Integration and support | Medium | Only deep, support-heavy, compliance-aware, enterprise-grade integrations—not generic wrappers around public APIs | | Support (SLA-backed) | — | Commit paid | Expertise | Priority response; guarantees | Low | Standard commercial | | Onboarding | — | Commit paid | Expertise | Implementation reviews; success plans | Low | Reduces risk and time-to-value | | Certification | — | Commit paid | Expertise | Operator/developer certification | Low | Trust and hiring signal | | Consulting / custom engineering | — | Commit paid | Expertise | Custom integration, architecture | Low | Expert labor | | Premium capability catalog / extended tool API | Extended verbs, proprietary adapters beyond open registry | Defer | Enterprise | Too early; validate core packages first | Medium | Revisit when demand exists | | Premium training / certified curriculum | Expanded golden, industry curriculum, certified model packs | Defer | Enterprise / Expertise | Too early; need product/support maturity | Medium | Revisit when demand clear | | Thin dashboard (UI over open scripts only) | — | Defer | — | Too easy to clone; weak product | High | Do not position as premium | | Standalone “run this script in cloud” | — | Defer | — | Shallow wrapper; weak | High | Only sell as part of governance/support | | Weak connector pack (thin API wrapper only) | — | Defer | — | Easy to recreate | High | Only ship connectors with depth | | Enterprise tool API demand | — | Discuss later | Enterprise | Revisit when we have customer signal | — | Trigger: first credible enterprise ask for premium Tool API | | Industry-specific curriculum demand | — | Discuss later | Enterprise / Expertise | Revisit when we have demand | — | Trigger: repeated onboarding/training demand or first real buyer for certified/vertical curriculum |

C. Commit open

We commit the following as open now. These surfaces will remain free and open for trust, adoption, and language legitimacy.

Spec and grammar: AINL language spec, grammar references, core and profile docs (docs/AINL_SPEC.md, docs/language/grammar.md, docs/language/AINL_CORE_AND_MODULES.md, docs/reference/AINL_V0_9_PROFILE.md).
Compiler: Parser, compiler, canonical IR generation (compiler_v2.py, compiler_grammar.py, grammar_constraint.py, grammar_priors.py).
Runtime: Reference runtime engine and execution semantics (runtime/engine.py, runtime/compat.py, runtime/values.py, SEMANTICS.md, docs/RUNTIME_COMPILER_CONTRACT.md).
Schemas: IR schema, graph schema (docs/reference/IR_SCHEMA.md, docs/reference/GRAPH_SCHEMA.md).
Baseline validation: Validation and conformance tooling and tests (scripts/validate_ainl.py, conformance tests, docs/CONFORMANCE.md).
Core docs and examples: Architecture, install, conformance, DOCS_INDEX, README; baseline examples and golden curriculum (examples/golden/*.ainl, docs/EXAMPLE_SUPPORT_MATRIX.md).
Onboarding and preflight: Bot onboarding and implementation preflight (docs/BOT_ONBOARDING.md, docs/OPENCLAW_IMPLEMENTATION_PREFLIGHT.md, tooling/bot_bootstrap.json).
Basic policy validator: IR-level policy check (tooling/policy_validator.py).
Health envelope schema: Standardized health envelope (docs/operations/STANDARDIZED_HEALTH_ENVELOPE.md).
Core adapter discovery / registry / manifest: Adapter registry and manifest for core adapters (docs/reference/ADAPTER_REGISTRY.md, tooling/adapter_manifest.json, core entries in ADAPTER_REGISTRY.json).
Capability registry (open subset): Schema and canonical-lane capability entries (tooling/capabilities.json, capabilities.schema.json, docs/reference/CAPABILITY_REGISTRY.md).
Baseline eval harness and report schemas: Concepts and open report schemas for reproducibility.
CLI fundamentals: Validate, run, basic emit (cli/main.py and related entrypoints).
Core runtime adapters: http, sqlite, fs, tools, wasm, base, replay (runtime/adapters/).
Extension adapters and examples (reference): OpenClaw integration, agent adapter, examples (adapters/openclaw_integration.py, adapters/agent.py, examples/openclaw/).
Graph/IR and capability scripts: Graph API, export, rewrite, ir_canonical, ir_compact, step_focus, trace_focus; capability report and filter scripts; memory tooling (retention, validate, import/export); operator-only audit script; coordination validator and mailbox validation.
Autonomous ops examples and run scripts: Reference monitors and patterns (examples/autonomous_ops/, demo/*.lang, scripts/run_*.py, docs/operations/AUTONOMOUS_OPS_MONITORS.md).
Memory contract: docs/adapters/MEMORY_CONTRACT.md.
Tool API v2 base projection: Open capability subset (scripts/gen_tool_api_v2_tools.py, tooling/tool_api_v2.tools.json).

D. Commit paid

We commit the following commercial pillars now. Paid value is in operational guarantees, governance/compliance, managed service, and expert support—not in thin wrappers or easily cloned utilities.

Managed

Hosted AINL runtime / managed ops platform: Run AINL at scale; scheduling; uptime; reduces operator labor.
Managed alignment pipeline: We run sweep, eval gate, trends, run health; SLA.
Managed dashboard / ops platform: Only when it provides aggregation, trends, regression visibility, alerts, governance integration, or SLA-backed operational value—not just UI over open outputs.

Enterprise

Governance / audit / compliance suite: Operator-only audit, capability governance, coordination validation, memory retention/hygiene, policy packs—packaged with integration and support.
Deployment kits: SSO, K8s, terraform, hardening playbooks; deploy in your VPC.
Policy packs: Curated, industry/compliance policies built on open validator; integration and support.
Supported ops monitor pack: Curated autonomous ops + runbooks + health envelope + updates + SLA (examples stay open; support and guarantees = paid).
Premium connector packs: Only deep, support-heavy, compliance-aware, enterprise-grade integrations—not generic wrappers around public APIs.

Expertise

Support: SLA-backed; priority response; guarantees.
Onboarding: Implementation reviews; success plans.
Certification: Operator/developer certification.
Consulting: Custom integration; architecture; custom engineering.

E. Defer

We do not decide now on the following. Revisit when we have product/support maturity or clear demand.

Premium capability catalog / extended tool API: Extended verbs, proprietary adapters, curated sets beyond open registry. Too early; validate core packages first.
Premium training / certified curriculum: Expanded golden, industry curriculum, certified model packs. Need product and support maturity before committing.
Thin offerings we will not position as premium: Thin dashboard (UI over open scripts only); standalone “run this script in the cloud” with no policy packs or integration; weak connector pack (thin API wrapper only). These are too easy to clone and do not become our paid strategy. Do not let them drift into product plans later.

F. Discuss later

Revisit when we have demand or execution reality (explicit triggers below):

Enterprise tool API: Revisit after first credible enterprise ask for premium Tool API or capability catalog.
Industry-specific curriculum: Revisit after repeated onboarding/training demand or first real buyer asking for certified/vertical curriculum.

These are not committed; they are on the list for future discussion.

G. Guardrails

Hard rules for boundary and product decisions:

Do not paywall language legitimacy. Spec, compiler, runtime, schemas, baseline validation, core docs/examples, onboarding/preflight, basic policy validator, health envelope schema, and core adapter discovery stay open.
Do not sell thin wrappers as premium strategy. Do not position a shallow UI over open scripts, or “we run this script for you” with no policy packs/integration/SLA, as a standalone premium product.
Do not make the open core useless. The free layer must be sufficient for real workflows (run, validate, core adapters, examples). Paid = convenience, scale, governance, support.
Do not surprise-relicense. Do not move previously open material to paid without announcement and documented rationale.
Do not blur open vs paid. Label clearly (Open Core / Commercial / Planned). Keep this decision sheet and docs/OPEN_CORE_BOUNDARY_MAP.md aligned.

H. Final recommended packaging

| Package | Contents | Role | |---------|----------|------| | Open Core | Spec, compiler, runtime, core adapters, schemas, baseline validation, core docs/examples, onboarding/preflight, basic policy validator, health envelope schema, core adapter registry and open capability subset, baseline eval/report schemas, CLI, graph/IR and capability scripts, memory/coordination/audit scripts, extension adapters and autonomous ops as reference. | Free, open. | | Managed Platform | Hosted AINL runtime; managed alignment pipeline; managed dashboard/ops (with meaningful operational value). Operational guarantees and SLA. | Paid. | | Enterprise Governance | Governance/audit suite (audit, capability governance, coordination validation, memory hygiene, policy packs); deployment kits; premium connector packs (when deep); supported ops monitor pack. | Paid. | | Expert Services | SLA-backed support; onboarding and success plans; certification; consulting and custom engineering. | Paid. |

Open Core is the free layer. Managed Platform, Enterprise Governance, and Expert Services are commercial. We do not overcommit to weak or easily cloned products; we bias paid strategy toward operational guarantees, governance/compliance, managed service, and expert support.

Boundary approval and lock

The open-core boundary is approved with minor refinements already incorporated into this decision sheet. The boundary is now considered locked for planning purposes. Further changes should only happen in response to concrete product, customer, or operational evidence.

Final signoff choices and notes are recorded in docs/OPEN_CORE_REVIEW_CHECKLIST.md (Section 4. Final signoff notes; Section 5. Boundary lock). For execution planning on first deliverables per pillar, see docs/OPEN_CORE_FIRST_DELIVERABLES.md.

This document is the final decision-layer planning artifact. No runtime or compiler semantics, licenses, or product code have been changed. For strategic context, see docs/MONETIZATION_OPEN_CORE_AUDIT.md. For the full product matrix and decision prompts, see docs/PRODUCT_MATRIX_OPEN_CORE.md.

Open-Core Decision Sheet

Source: Distilled from docs/MONETIZATION_OPEN_CORE_AUDIT.md and docs/PRODUCT_MATRIX_OPEN_CORE.md.

Positioning:

AINL is the open language for agent workflows. We sell the operational layer: governance, visibility, managed runtime, deployment, and support.

Strategic rule:

Language legitimacy stays open.
Operator labor / governance / convenience gets monetized.

Paid buckets: Managed | Enterprise | Expertise. We do not position thin, easily cloned utilities as premium products.

A. Decision summary

B. Decision table

C. Commit open

We commit the following as open now. These surfaces will remain free and open for trust, adoption, and language legitimacy.

Spec and grammar: AINL language spec, grammar references, core and profile docs (docs/AINL_SPEC.md, docs/language/grammar.md, docs/language/AINL_CORE_AND_MODULES.md, docs/reference/AINL_V0_9_PROFILE.md).
Compiler: Parser, compiler, canonical IR generation (compiler_v2.py, compiler_grammar.py, grammar_constraint.py, grammar_priors.py).
Runtime: Reference runtime engine and execution semantics (runtime/engine.py, runtime/compat.py, runtime/values.py, SEMANTICS.md, docs/RUNTIME_COMPILER_CONTRACT.md).
Schemas: IR schema, graph schema (docs/reference/IR_SCHEMA.md, docs/reference/GRAPH_SCHEMA.md).
Baseline validation: Validation and conformance tooling and tests (scripts/validate_ainl.py, conformance tests, docs/CONFORMANCE.md).
Core docs and examples: Architecture, install, conformance, DOCS_INDEX, README; baseline examples and golden curriculum (examples/golden/*.ainl, docs/EXAMPLE_SUPPORT_MATRIX.md).
Onboarding and preflight: Bot onboarding and implementation preflight (docs/BOT_ONBOARDING.md, docs/OPENCLAW_IMPLEMENTATION_PREFLIGHT.md, tooling/bot_bootstrap.json).
Basic policy validator: IR-level policy check (tooling/policy_validator.py).
Health envelope schema: Standardized health envelope (docs/operations/STANDARDIZED_HEALTH_ENVELOPE.md).
Core adapter discovery / registry / manifest: Adapter registry and manifest for core adapters (docs/reference/ADAPTER_REGISTRY.md, tooling/adapter_manifest.json, core entries in ADAPTER_REGISTRY.json).
Capability registry (open subset): Schema and canonical-lane capability entries (tooling/capabilities.json, capabilities.schema.json, docs/reference/CAPABILITY_REGISTRY.md).
Baseline eval harness and report schemas: Concepts and open report schemas for reproducibility.
CLI fundamentals: Validate, run, basic emit (cli/main.py and related entrypoints).
Core runtime adapters: http, sqlite, fs, tools, wasm, base, replay (runtime/adapters/).
Extension adapters and examples (reference): OpenClaw integration, agent adapter, examples (adapters/openclaw_integration.py, adapters/agent.py, examples/openclaw/).
Graph/IR and capability scripts: Graph API, export, rewrite, ir_canonical, ir_compact, step_focus, trace_focus; capability report and filter scripts; memory tooling (retention, validate, import/export); operator-only audit script; coordination validator and mailbox validation.
Autonomous ops examples and run scripts: Reference monitors and patterns (examples/autonomous_ops/, demo/*.lang, scripts/run_*.py, docs/operations/AUTONOMOUS_OPS_MONITORS.md).
Memory contract: docs/adapters/MEMORY_CONTRACT.md.
Tool API v2 base projection: Open capability subset (scripts/gen_tool_api_v2_tools.py, tooling/tool_api_v2.tools.json).

D. Commit paid

Managed

Hosted AINL runtime / managed ops platform: Run AINL at scale; scheduling; uptime; reduces operator labor.
Managed alignment pipeline: We run sweep, eval gate, trends, run health; SLA.
Managed dashboard / ops platform: Only when it provides aggregation, trends, regression visibility, alerts, governance integration, or SLA-backed operational value—not just UI over open outputs.

Enterprise

Governance / audit / compliance suite: Operator-only audit, capability governance, coordination validation, memory retention/hygiene, policy packs—packaged with integration and support.
Deployment kits: SSO, K8s, terraform, hardening playbooks; deploy in your VPC.
Policy packs: Curated, industry/compliance policies built on open validator; integration and support.
Supported ops monitor pack: Curated autonomous ops + runbooks + health envelope + updates + SLA (examples stay open; support and guarantees = paid).
Premium connector packs: Only deep, support-heavy, compliance-aware, enterprise-grade integrations—not generic wrappers around public APIs.

Expertise

Support: SLA-backed; priority response; guarantees.
Onboarding: Implementation reviews; success plans.
Certification: Operator/developer certification.
Consulting: Custom integration; architecture; custom engineering.

E. Defer

We do not decide now on the following. Revisit when we have product/support maturity or clear demand.

Premium capability catalog / extended tool API: Extended verbs, proprietary adapters, curated sets beyond open registry. Too early; validate core packages first.
Premium training / certified curriculum: Expanded golden, industry curriculum, certified model packs. Need product and support maturity before committing.
Thin offerings we will not position as premium: Thin dashboard (UI over open scripts only); standalone “run this script in the cloud” with no policy packs or integration; weak connector pack (thin API wrapper only). These are too easy to clone and do not become our paid strategy. Do not let them drift into product plans later.

F. Discuss later

Revisit when we have demand or execution reality (explicit triggers below):

Enterprise tool API: Revisit after first credible enterprise ask for premium Tool API or capability catalog.
Industry-specific curriculum: Revisit after repeated onboarding/training demand or first real buyer asking for certified/vertical curriculum.

These are not committed; they are on the list for future discussion.

G. Guardrails

Hard rules for boundary and product decisions:

Do not paywall language legitimacy. Spec, compiler, runtime, schemas, baseline validation, core docs/examples, onboarding/preflight, basic policy validator, health envelope schema, and core adapter discovery stay open.
Do not sell thin wrappers as premium strategy. Do not position a shallow UI over open scripts, or “we run this script for you” with no policy packs/integration/SLA, as a standalone premium product.
Do not make the open core useless. The free layer must be sufficient for real workflows (run, validate, core adapters, examples). Paid = convenience, scale, governance, support.
Do not surprise-relicense. Do not move previously open material to paid without announcement and documented rationale.
Do not blur open vs paid. Label clearly (Open Core / Commercial / Planned). Keep this decision sheet and docs/OPEN_CORE_BOUNDARY_MAP.md aligned.

H. Final recommended packaging

Boundary approval and lock

Open-Core Decision Sheet

Source: Distilled from docs/MONETIZATION_OPEN_CORE_AUDIT.md and docs/PRODUCT_MATRIX_OPEN_CORE.md.

Positioning:

AINL is the open language for agent workflows. We sell the operational layer: governance, visibility, managed runtime, deployment, and support.

Strategic rule:

Language legitimacy stays open.
Operator labor / governance / convenience gets monetized.

Paid buckets: Managed | Enterprise | Expertise. We do not position thin, easily cloned utilities as premium products.

A. Decision summary

B. Decision table

C. Commit open

We commit the following as open now. These surfaces will remain free and open for trust, adoption, and language legitimacy.

Spec and grammar: AINL language spec, grammar references, core and profile docs (docs/AINL_SPEC.md, docs/language/grammar.md, docs/language/AINL_CORE_AND_MODULES.md, docs/reference/AINL_V0_9_PROFILE.md).
Compiler: Parser, compiler, canonical IR generation (compiler_v2.py, compiler_grammar.py, grammar_constraint.py, grammar_priors.py).
Runtime: Reference runtime engine and execution semantics (runtime/engine.py, runtime/compat.py, runtime/values.py, SEMANTICS.md, docs/RUNTIME_COMPILER_CONTRACT.md).
Schemas: IR schema, graph schema (docs/reference/IR_SCHEMA.md, docs/reference/GRAPH_SCHEMA.md).
Baseline validation: Validation and conformance tooling and tests (scripts/validate_ainl.py, conformance tests, docs/CONFORMANCE.md).
Core docs and examples: Architecture, install, conformance, DOCS_INDEX, README; baseline examples and golden curriculum (examples/golden/*.ainl, docs/EXAMPLE_SUPPORT_MATRIX.md).
Onboarding and preflight: Bot onboarding and implementation preflight (docs/BOT_ONBOARDING.md, docs/OPENCLAW_IMPLEMENTATION_PREFLIGHT.md, tooling/bot_bootstrap.json).
Basic policy validator: IR-level policy check (tooling/policy_validator.py).
Health envelope schema: Standardized health envelope (docs/operations/STANDARDIZED_HEALTH_ENVELOPE.md).
Core adapter discovery / registry / manifest: Adapter registry and manifest for core adapters (docs/reference/ADAPTER_REGISTRY.md, tooling/adapter_manifest.json, core entries in ADAPTER_REGISTRY.json).
Capability registry (open subset): Schema and canonical-lane capability entries (tooling/capabilities.json, capabilities.schema.json, docs/reference/CAPABILITY_REGISTRY.md).
Baseline eval harness and report schemas: Concepts and open report schemas for reproducibility.
CLI fundamentals: Validate, run, basic emit (cli/main.py and related entrypoints).
Core runtime adapters: http, sqlite, fs, tools, wasm, base, replay (runtime/adapters/).
Extension adapters and examples (reference): OpenClaw integration, agent adapter, examples (adapters/openclaw_integration.py, adapters/agent.py, examples/openclaw/).
Graph/IR and capability scripts: Graph API, export, rewrite, ir_canonical, ir_compact, step_focus, trace_focus; capability report and filter scripts; memory tooling (retention, validate, import/export); operator-only audit script; coordination validator and mailbox validation.
Autonomous ops examples and run scripts: Reference monitors and patterns (examples/autonomous_ops/, demo/*.lang, scripts/run_*.py, docs/operations/AUTONOMOUS_OPS_MONITORS.md).
Memory contract: docs/adapters/MEMORY_CONTRACT.md.
Tool API v2 base projection: Open capability subset (scripts/gen_tool_api_v2_tools.py, tooling/tool_api_v2.tools.json).

D. Commit paid

Managed

Hosted AINL runtime / managed ops platform: Run AINL at scale; scheduling; uptime; reduces operator labor.
Managed alignment pipeline: We run sweep, eval gate, trends, run health; SLA.
Managed dashboard / ops platform: Only when it provides aggregation, trends, regression visibility, alerts, governance integration, or SLA-backed operational value—not just UI over open outputs.

Enterprise

Governance / audit / compliance suite: Operator-only audit, capability governance, coordination validation, memory retention/hygiene, policy packs—packaged with integration and support.
Deployment kits: SSO, K8s, terraform, hardening playbooks; deploy in your VPC.
Policy packs: Curated, industry/compliance policies built on open validator; integration and support.
Supported ops monitor pack: Curated autonomous ops + runbooks + health envelope + updates + SLA (examples stay open; support and guarantees = paid).
Premium connector packs: Only deep, support-heavy, compliance-aware, enterprise-grade integrations—not generic wrappers around public APIs.

Expertise

Support: SLA-backed; priority response; guarantees.
Onboarding: Implementation reviews; success plans.
Certification: Operator/developer certification.
Consulting: Custom integration; architecture; custom engineering.

E. Defer

We do not decide now on the following. Revisit when we have product/support maturity or clear demand.

Premium capability catalog / extended tool API: Extended verbs, proprietary adapters, curated sets beyond open registry. Too early; validate core packages first.
Premium training / certified curriculum: Expanded golden, industry curriculum, certified model packs. Need product and support maturity before committing.
Thin offerings we will not position as premium: Thin dashboard (UI over open scripts only); standalone “run this script in the cloud” with no policy packs or integration; weak connector pack (thin API wrapper only). These are too easy to clone and do not become our paid strategy. Do not let them drift into product plans later.

F. Discuss later

Revisit when we have demand or execution reality (explicit triggers below):

Enterprise tool API: Revisit after first credible enterprise ask for premium Tool API or capability catalog.
Industry-specific curriculum: Revisit after repeated onboarding/training demand or first real buyer asking for certified/vertical curriculum.

These are not committed; they are on the list for future discussion.

G. Guardrails

Hard rules for boundary and product decisions:

Do not paywall language legitimacy. Spec, compiler, runtime, schemas, baseline validation, core docs/examples, onboarding/preflight, basic policy validator, health envelope schema, and core adapter discovery stay open.
Do not sell thin wrappers as premium strategy. Do not position a shallow UI over open scripts, or “we run this script for you” with no policy packs/integration/SLA, as a standalone premium product.
Do not make the open core useless. The free layer must be sufficient for real workflows (run, validate, core adapters, examples). Paid = convenience, scale, governance, support.
Do not surprise-relicense. Do not move previously open material to paid without announcement and documented rationale.
Do not blur open vs paid. Label clearly (Open Core / Commercial / Planned). Keep this decision sheet and docs/OPEN_CORE_BOUNDARY_MAP.md aligned.