@neuroverseos/governance 0.3.0 → 0.3.1

package/dist/worlds/autoresearch.nv-world.md

@@ -1,230 +0,0 @@
----
-world_id: autoresearch
-name: Autoresearch Governance
-version: 1.0.0
-runtime_mode: SIMULATION
-default_profile: conservative
-alternative_profile: exploratory
----
-
-# Thesis
-
-Autonomous AI research loops must operate within structured governance: experiments are reproducible, metrics are tracked, compute budgets are enforced, and agents cannot drift beyond their declared research context. A research world without constraints produces noise, not knowledge.
-
-# Invariants
-
-- `experiments_must_be_reproducible` — Every experiment must log architecture, hyperparameters, dataset, and training config sufficient to reproduce results (structural, immutable)
-- `metrics_must_be_recorded` — Every training run must produce at least one evaluation metric; runs without metrics are invalid (structural, immutable)
-- `dataset_must_be_declared` — The dataset used for training and evaluation must be explicitly declared and never changed without governance approval (structural, immutable)
-- `goal_must_be_defined` — The optimization goal (metric + direction) must be defined before any experiment runs (structural, immutable)
-- `no_data_leakage` — Training data must never contaminate evaluation data; train/val/test splits must be fixed (structural, immutable)
-- `compute_budget_enforced` — Experiments must respect declared compute limits; exceeding budget halts the loop (structural, immutable)
-- `architecture_constraints_honored` — If the research context declares architectural constraints, experiments must satisfy them (prompt, immutable)
-
-# State
-
-## experiments_run
-- type: number
-- min: 0
-- max: 10000
-- step: 1
-- default: 0
-- label: Experiments Run
-- description: Total number of experiments completed in this research loop
-
-## best_metric_value
-- type: number
-- min: -1000
-- max: 1000
-- step: 0.01
-- default: 100
-- label: Best Metric Value
-- description: Best value achieved for the primary evaluation metric
-
-## keep_rate
-- type: number
-- min: 0
-- max: 100
-- step: 1
-- default: 0
-- label: Keep Rate
-- description: Percentage of experiments that improved upon the previous best result
-
-## compute_used_minutes
-- type: number
-- min: 0
-- max: 100000
-- step: 1
-- default: 0
-- label: Compute Used (minutes)
-- description: Total wall-clock training time consumed across all experiments
-
-## compute_budget_minutes
-- type: number
-- min: 0
-- max: 100000
-- step: 60
-- default: 1440
-- label: Compute Budget (minutes)
-- description: Maximum allowed wall-clock training time for the research loop
-
-## research_context_drift
-- type: number
-- min: 0
-- max: 100
-- step: 1
-- default: 0
-- label: Context Drift
-- description: Degree to which recent experiments have diverged from the declared research context. 0 = on-topic. 100 = unrelated.
-
-## metric_improvement_rate
-- type: number
-- min: 0
-- max: 100
-- step: 1
-- default: 0
-- label: Improvement Rate
-- description: Rate of metric improvement over the last 10 experiments. 0 = stagnant. 100 = rapid improvement.
-
-## failed_experiments
-- type: number
-- min: 0
-- max: 10000
-- step: 1
-- default: 0
-- label: Failed Experiments
-- description: Number of experiments that crashed, timed out, or produced no valid metrics
-
-# Assumptions
-
-## conservative
-- name: Conservative Research
-- description: Prioritize reproducibility and careful iteration. Small architectural changes per experiment. Strict compute limits. Reject experiments that drift from the research context.
-- iteration_style: incremental
-- drift_tolerance: low
-- compute_strictness: high
-- failure_tolerance: low
-
-## exploratory
-- name: Exploratory Research
-- description: Allow broader architectural exploration. Larger jumps between experiments. More lenient compute budget. Accept higher context drift if metrics improve.
-- iteration_style: explorative
-- drift_tolerance: moderate
-- compute_strictness: moderate
-- failure_tolerance: moderate
-
-# Rules
-
-## rule-001: Compute Budget Exhausted (structural)
-When compute budget is exceeded, the research loop must halt. No further experiments are allowed.
-
-When compute_used_minutes > compute_budget_minutes [state]
-Then research_viability *= 0.00
-Collapse: research_viability < 0.05
-
-> trigger: Compute usage exceeds declared budget — no training time remains.
-> rule: Unbounded compute makes research ungovernable. The budget is a hard constraint, not a suggestion.
-> shift: Research loop halts. Final results are reported. No new experiments start.
-> effect: Research viability set to zero. Loop terminated.
-
-## rule-002: High Failure Rate (degradation)
-Too many failed experiments indicate a systemic problem — bad code, misconfigured environment, or impossible architecture.
-
-When failed_experiments > 5 [state] AND experiments_run > 0 [state]
-Then research_viability *= 0.50
-
-> trigger: More than 5 experiments have failed — possible systemic issue.
-> rule: Failures consume compute without producing knowledge. High failure rates signal infrastructure problems, not research progress.
-> shift: Research viability degrades. Agent should investigate root cause before continuing.
-> effect: Research viability reduced to 50%.
-
-## rule-003: Context Drift Warning (degradation)
-Experiments diverging from the declared research context waste compute and produce irrelevant results.
-
-When research_context_drift > 40 [state]
-Then research_viability *= 0.60
-
-> trigger: Context drift above 40% — experiments are straying from the research topic.
-> rule: Governance exists to keep research focused. Agents exploring unrelated architectures are not contributing to the declared goal.
-> shift: Research viability degrades. Agent must return to the declared research context.
-> effect: Research viability reduced to 60%.
-
-## rule-004: Metric Stagnation (degradation)
-When experiments stop improving the primary metric, the research approach may need fundamental revision.
-
-When metric_improvement_rate < 5 [state] AND experiments_run > 10 [state]
-Then research_viability *= 0.70
-
-> trigger: Improvement rate below 5% after 10+ experiments — research may have plateaued.
-> rule: Stagnant metrics indicate diminishing returns from the current approach. The agent should consider a strategy change.
-> shift: Research viability degrades. Agent should try a substantially different approach or conclude the loop.
-> effect: Research viability reduced to 70%.
-
-## rule-005: Strong Progress (advantage)
-Consistent metric improvement validates the research approach and warrants continued investment.
-
-When metric_improvement_rate > 30 [state] AND keep_rate > 20 [state]
-Then research_viability *= 1.20
-
-> trigger: Improvement rate above 30% with keep rate above 20% — research is productive.
-> rule: Productive research should be encouraged. Strong metric trends indicate a promising research direction.
-> shift: Research viability improves. Continued experimentation is well-justified.
-> effect: Research viability boosted by 20%.
-
-## rule-006: No Metrics Recorded (structural)
-An experiment that produces no evaluation metrics is invalid and must not count as progress.
-
-When experiments_run > 0 [state] AND best_metric_value == 100 [state]
-Then research_viability *= 0.30
-Collapse: research_viability < 0.05
-
-> trigger: Experiments have run but no metric improvement from default — metrics may not be recording.
-> rule: Research without measurement is not research. Every experiment must produce at least one evaluation metric.
-> shift: Research viability drops sharply. Agent must fix metric recording before continuing.
-> effect: Research viability reduced to 30%.
-
-## rule-007: Efficient Compute Usage (advantage)
-High keep rate with low compute usage indicates efficient research methodology.
-
-When keep_rate > 30 [state] AND compute_used_minutes < compute_budget_minutes [state]
-Then research_viability *= 1.15
-
-> trigger: Keep rate above 30% with compute budget remaining — efficient experimentation.
-> rule: Efficient use of compute demonstrates disciplined research. Not every experiment needs to be expensive.
-> shift: Research viability improves. The research methodology is sustainable.
-> effect: Research viability boosted by 15%.
-
-# Gates
-
-- BREAKTHROUGH: research_viability >= 90
-- PRODUCTIVE: research_viability >= 60
-- ONGOING: research_viability >= 35
-- STRUGGLING: research_viability > 10
-- HALTED: research_viability <= 10
-
-# Outcomes
-
-## research_viability
-- type: number
-- range: 0-100
-- display: percentage
-- label: Research Viability
-- primary: true
-
-## best_metric_value
-- type: number
-- range: -1000-1000
-- display: decimal
-- label: Best Metric Value
-
-## keep_rate
-- type: number
-- range: 0-100
-- display: percentage
-- label: Keep Rate
-
-## experiments_run
-- type: number
-- range: 0-10000
-- display: integer
-- label: Experiments Run

package/dist/worlds/derivation-world.nv-world.md

@@ -1,278 +0,0 @@
----
-world_id: derivationworld
-name: DerivationWorld
-version: 1.0.0
-runtime_mode: synthesis
-default_profile: strict_synthesis
-alternative_profile: permissive_synthesis
----
-
-# Thesis
-
-AI-synthesized governance documents must be structurally valid, epistemically honest, and deterministically verifiable. A derived .nv-world.md is only legitimate if it satisfies the same parser constraints as a hand-authored world, distinguishes declared facts from inferred claims, and never introduces governance domains beyond the source material.
-
-# Invariants
-
-- `output_must_be_valid_nv_world` — Synthesized output must parse successfully under parseWorldMarkdown with zero errors (prompt, immutable)
-- `must_include_required_sections` — Output must contain Thesis, Invariants, State, Rules, Gates, and Outcomes sections (prompt, immutable)
-- `must_distinguish_declared_vs_inferred` — Invariants derived from explicit source statements must be marked structural; those inferred by the model must be marked operational (prompt, immutable)
-- `must_not_invent_external_domains` — All state variables, rules, and invariants must trace to concepts present in the input markdown (prompt, immutable)
-- `invariants_must_be_enforceable_or_marked` — Every invariant must be structurally enforceable via rules, or explicitly tagged as non-enforceable with rationale (prompt, immutable)
-- `no_json_output` — Output must be .nv-world.md markdown only, never JSON (prompt, immutable)
-- `no_extra_commentary` — Output must contain only the .nv-world.md document, no preamble, explanation, or trailing commentary (prompt, immutable)
-- `frontmatter_must_be_complete` — Output frontmatter must include world_id, name, and version fields (prompt, immutable)
-- `rules_must_have_triggers_and_effects` — Every rule must include a When trigger line and a Then effect line (prompt, immutable)
-- `gate_thresholds_must_be_ordered` — Gate thresholds must be monotonically decreasing from best to worst status (prompt, immutable)
-
-# State
-
-## source_section_count
-- type: number
-- min: 0
-- max: 100
-- step: 1
-- default: 5
-- label: Source Section Count
-- description: Number of distinct sections or files in the input markdown. More sections generally means richer synthesis material.
-
-## source_token_estimate
-- type: number
-- min: 0
-- max: 200000
-- step: 100
-- default: 2000
-- label: Source Token Estimate
-- description: Approximate token count of concatenated input. Determines whether context window constraints may truncate material.
-
-## declared_concept_count
-- type: number
-- min: 0
-- max: 200
-- step: 1
-- default: 10
-- label: Declared Concept Count
-- description: Number of distinct governance concepts explicitly named in source material. Drives state variable and rule generation.
-
-## concept_specificity
-- type: number
-- min: 0
-- max: 100
-- default: 50
-- label: Concept Specificity
-- description: How precisely the source material defines its governance concepts. 0 = vague aspirations. 100 = precise structural claims with measurable criteria.
-
-## domain_coherence
-- type: number
-- min: 0
-- max: 100
-- default: 60
-- label: Domain Coherence
-- description: How well source sections relate to a single governance domain. Low coherence indicates conflicting or unrelated source material.
-
-## synthesis_fidelity
-- type: number
-- min: 0.00
-- max: 1.00
-- step: 0.01
-- default: 0.70
-- label: Synthesis Fidelity
-- description: Measure of how faithfully the derived world represents the source material. Primary outcome metric.
-
-## structural_completeness
-- type: number
-- min: 0
-- max: 100
-- default: 60
-- label: Structural Completeness
-- description: Percentage of required .nv-world.md sections that contain meaningful content rather than stubs.
-
-## epistemic_honesty
-- type: number
-- min: 0
-- max: 100
-- default: 70
-- label: Epistemic Honesty
-- description: Degree to which the output correctly distinguishes source-declared constraints from model-inferred constraints. 0 = everything claimed as declared. 100 = perfect attribution.
-
-## invention_ratio
-- type: number
-- min: 0.00
-- max: 1.00
-- step: 0.01
-- default: 0.10
-- label: Invention Ratio
-- description: Fraction of output concepts that have no traceable origin in the source material. Should be near zero. Above 0.30 indicates hallucination.
-
-# Assumptions
-
-## strict_synthesis
-- name: Strict Synthesis
-- description: Conservative derivation that only produces governance elements with clear source basis. Prefers omission over invention. Marks all inferred elements as operational.
-- invention_tolerance: minimal
-- attribution_mode: strict
-- completeness_priority: low
-- fidelity_priority: high
-
-## permissive_synthesis
-- name: Permissive Synthesis
-- description: Broader derivation that fills structural gaps with reasonable inferences. Produces more complete worlds but with higher invention ratio. All inferences are still marked operational.
-- invention_tolerance: moderate
-- attribution_mode: standard
-- completeness_priority: high
-- fidelity_priority: moderate
-
-# Rules
-
-## rule-001: Empty Source Rejection (structural)
-Synthesis from empty or trivially short input cannot produce meaningful governance. The derivation must fail rather than fabricate.
-
-When source_section_count < 1 [state]
-Then synthesis_fidelity *= 0.00
-Collapse: synthesis_fidelity < 0.05
-
-> trigger: Source input contains no sections — nothing to derive from.
-> rule: A world cannot be synthesized from nothing. Empty input must produce a clear failure, not a fabricated world.
-> shift: Derivation halts. No output file is written.
-> effect: Synthesis fidelity set to zero. Derivation rejected.
-
-## rule-002: Sparse Source Warning (degradation)
-Minimal source material limits the quality of derived governance. Output will be structurally thin.
-
-When source_section_count < 3 [state] AND source_token_estimate < 500 [state]
-Then synthesis_fidelity *= 0.50, structural_completeness *= 0.60
-
-> trigger: Source has fewer than 3 sections and under 500 tokens — sparse material.
-> rule: Sparse input yields sparse governance. The model cannot reliably infer structure from fragments.
-> shift: Output quality degrades. State variables and rules will be minimal.
-> effect: Synthesis fidelity reduced to 50%. Structural completeness reduced to 60%.
-
-## rule-003: Concept Vagueness Penalty (degradation)
-Source material with low concept specificity produces invariants and rules that are aspirational rather than structural.
-
-When concept_specificity < 25 [state]
-Then synthesis_fidelity *= 0.60, epistemic_honesty *= 0.70
-
-> trigger: Source concept specificity below 25% — governance concepts are vague.
-> rule: Vague concepts cannot produce structural invariants. The model must either invent specificity or produce unenforceable constraints.
-> shift: Output invariants trend toward aspiration. Rules lack deterministic triggers.
-> effect: Synthesis fidelity reduced to 60%. Epistemic honesty reduced to 70%.
-
-## rule-004: Domain Incoherence Penalty (degradation)
-Source material spanning unrelated domains produces a world with conflicting governance logic.
-
-When domain_coherence < 30 [state]
-Then synthesis_fidelity *= 0.55
-
-> trigger: Domain coherence below 30% — source material is internally contradictory or covers unrelated domains.
-> rule: A single .nv-world.md should govern a coherent domain. Mixed domains produce conflicting rules and meaningless invariants.
-> shift: Output becomes structurally confused. State variables may not relate to each other.
-> effect: Synthesis fidelity reduced to 55%.
-
-## rule-005: Invention Threshold Breach (structural)
-Excessive invention without source basis constitutes fabrication, not derivation.
-
-When invention_ratio > 0.30 [state]
-Then synthesis_fidelity *= 0.30, epistemic_honesty *= 0.40
-Collapse: synthesis_fidelity < 0.05
-
-> trigger: Invention ratio exceeds 30% — more than a third of output has no source basis.
-> rule: Derivation must be grounded. A world that is mostly invented does not represent the user's governance intent.
-> shift: Output crosses from synthesis to hallucination. Fidelity drops below usable threshold.
-> effect: Synthesis fidelity reduced to 30%. Epistemic honesty reduced to 40%.
-
-## rule-006: High Fidelity Source (advantage)
-Rich, specific, coherent source material enables high-quality derivation.
-
-When concept_specificity > 70 [state] AND domain_coherence > 70 [state] AND declared_concept_count > 8 [state]
-Then synthesis_fidelity *= 1.20, structural_completeness *= 1.15
-
-> trigger: High concept specificity, strong domain coherence, and rich concept count.
-> rule: Quality source material produces quality governance. The model has enough structure to derive rather than invent.
-> shift: Output is well-grounded. Most invariants and rules trace directly to source.
-> effect: Synthesis fidelity boosted by 20%. Structural completeness boosted by 15%.
-
-## rule-007: Structural Completeness Gate (degradation)
-A derived world missing critical sections is not usable regardless of quality in present sections.
-
-When structural_completeness < 40 [state]
-Then synthesis_fidelity *= 0.50
-
-> trigger: Structural completeness below 40% — too many required sections are empty or stub.
-> rule: A partial world is not a valid world. Missing sections mean missing governance.
-> shift: The output may parse but cannot function as meaningful governance.
-> effect: Synthesis fidelity reduced to 50%.
-
-## rule-008: Epistemic Honesty Reward (advantage)
-Correct attribution of declared versus inferred constraints makes output trustworthy and auditable.
-
-When epistemic_honesty > 80 [state]
-Then synthesis_fidelity *= 1.10
-
-> trigger: Epistemic honesty above 80% — model correctly attributes constraint origins.
-> rule: Honest attribution makes governance auditable. Users can verify which constraints they declared versus which the model suggested.
-> shift: Output gains trust. Declared constraints can be relied upon; inferred ones can be reviewed.
-> effect: Synthesis fidelity boosted by 10%.
-
-## rule-009: Context Window Overflow Risk (degradation)
-Extremely large source material risks truncation and missed governance concepts.
-
-When source_token_estimate > 100000 [state]
-Then synthesis_fidelity *= 0.75
-
-> trigger: Source material exceeds 100k tokens — likely to be truncated.
-> rule: Truncated input means incomplete synthesis. The model may miss governance concepts that appear late in the concatenation.
-> shift: Output may be partial. Critical sections from later source files may be absent.
-> effect: Synthesis fidelity reduced to 75% due to truncation risk.
-
-## rule-010: Derivation Coherence Reward (advantage)
-Aligned quality metrics across fidelity, honesty, and invention produce a genuine governance document.
-
-When synthesis_fidelity > 0.80 [state] AND epistemic_honesty > 75 [state] AND invention_ratio < 0.15 [state]
-Then synthesis_fidelity *= 1.15
-Collapse: synthesis_fidelity < 0.05
-
-> trigger: Synthesis fidelity above 80%, epistemic honesty above 75%, and invention ratio below 15%.
-> rule: Coherent derivation across all metrics indicates a faithful, usable governance document.
-> shift: The derived world moves from draft to production-quality. Suitable for bootstrap and validation.
-> effect: Synthesis fidelity boosted by 15%. Derivation coherence achieved.
-
-# Gates
-
-- FAITHFUL: synthesis_fidelity >= 0.85
-- USABLE: synthesis_fidelity >= 0.60
-- REVIEWABLE: synthesis_fidelity >= 0.40
-- SUSPECT: synthesis_fidelity > 0.15
-- DERIVATION_REJECTED: synthesis_fidelity <= 0.15
-
-# Outcomes
-
-## synthesis_fidelity
-- type: number
-- range: 0-1
-- display: percentage
-- label: Synthesis Fidelity
-- primary: true
-
-## structural_completeness
-- type: number
-- range: 0-100
-- display: percentage
-- label: Structural Completeness
-
-## epistemic_honesty
-- type: number
-- range: 0-100
-- display: percentage
-- label: Epistemic Honesty
-
-## invention_ratio
-- type: number
-- range: 0-1
-- display: percentage
-- label: Invention Ratio
-- assignment: external
-
-## derivation_status
-- type: enum
-- label: Derivation Status
-- assignment: external