Mantic Early Warning System - Universal Manifest

name: mantic-early-warning
version: 1.4.1
description: Cross-domain anomaly and opportunity detection using 4-layer hierarchical analysis
author: Mantic Framework
license: Elastic License 2.0 (Source-Available) / Commercial licenses available
models: [claude, kimi, gemini, openai, ollama]

---

Core Formula (IMMUTABLE)

M = (sum(W * L * I)) * f(t) / k_n

Where:

• M: Final mantic score (anomaly/opportunity intensity)
• W: Array of 4 weights (must sum to 1)
• L: Array of 4 layer values (0-1 range)
• I: Array of 4 interaction coefficients (0-1)
• f(t): Temporal kernel value (default 1.0)
• k_n: Normalization constant (default 1.0)

Layers: [micro, meso, macro, meta] mapped to domain-specific inputs

---

Temporal Kernels

Models can pre-compute f_time using compute_temporal_kernel() before calling any detect() tool.
This allows time-aware scoring — signals decay, grow, saturate, or oscillate depending on the mode.

Usage

from mantic_thinking.core.mantic_kernel import compute_temporal_kernel

# Compute temporal modifier, then pass to any tool
f_time = compute_temporal_kernel(t=5, n=1.0, alpha=0.1, kernel_type="exponential")
result = detect(..., f_time=f_time)

Parameters

• t: Time delta (0 = now, positive = future, negative = past)
• n: Novelty — how unusual the pattern is (>0 amplifies, <0 attenuates)
• alpha: Sensitivity — how reactive to novelty over time
• kernel_type: Which temporal model to use

Available Modes

Mode	When to Use	Example Domain
exponential	Viral spread, cascade growth/decay	Social narrative rupture
logistic	Saturation, carrying capacity	Legal precedent drift
s_curve	Adoption curves, learning onset	Healthcare therapeutic windows
power_law	Heavy-tailed, long-memory dynamics	Climate maladaptation
oscillatory	Seasonal, cyclical patterns	Finance regime conflicts
memory	Decaying influence of past events	Military strategic initiative
linear	Simple linear decay	General purpose

Mode-Specific Parameters

• s_curve: t0 — inflection point (default 0.0)
• power_law: exponent — power coefficient (default 1.0)
• oscillatory: frequency — oscillation frequency (default 1.0)
• memory: memory_strength — initial memory weight (default 1.0)

Warning: The exponential mode with positive n and alpha produces growth, not decay.
Use n=-1 for decay behavior (matching old decay_rate parameter).

Note: power_law requires t >= -1. Values below this are clamped.

---

Tool Suites

The Mantic Framework provides 17 tools across 8 domains, with two complementary detection modes plus a generic endpoint:

Friction Suite (8 tools)

Logic: Detects divergences and risks (per-tool thresholds; see each tool's DEFAULT_THRESHOLDS)
Use: Risk assessment, anomaly detection, bottleneck identification
Output: Alerts, warnings, risk ratings

Emergence Suite (8 tools)

Logic: Detects alignments and opportunities (per-tool thresholds; see each tool's DEFAULT_THRESHOLDS)
Use: Optimal timing, high-leverage interventions, window detection
Output: Window detected, recommendations, timing guidance

---

Friction Tools (Divergence Detection)

1. healthcare_phenotype_genotype

Name: Healthcare: Phenotype-Genotype Mismatch Detector
Description: Detects when genomic risk doesn't match phenotypic presentation, indicating environmental buffering or psychosocial resilience

Input Layers:

Layer	Description	Type	Required
phenotypic	Current symptoms/vitals (0-1)	float	yes
genomic	Genetic risk score (0-1)	float	yes
environmental	Exposure load (0-1)	float	yes
psychosocial	Stress/resilience (0-1)	float	yes

Output:

{
  "alert": "string - detection message",
  "severity": "float 0-1",
  "buffering_layer": "string - which layer is buffering",
  "m_score": "float - final mantic score",
  "spatial_component": "float - S value",
  "layer_attribution": "object - percentage contribution per layer"
}

Trigger Threshold: 0.4

---

2. finance_regime_conflict

Name: Finance: Regime Conflict Monitor
Description: Spots when technical price action contradicts fundamentals, flow, or risk signals

Input Layers:

Layer	Description	Type	Range
technical	Price action signals	float	0-1
macro	Fundamental indicators	float	0-1
flow	Capital flow direction	float	[-1, 1]
risk	Risk appetite metrics	float	0-1

Output:

{
  "alert": "string",
  "conflict_type": "string - type of regime conflict",
  "confidence": "float 0-1",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

3. cyber_attribution_resolver

Name: Cybersecurity: Attribution Uncertainty Resolver
Description: Scores confidence when technical sophistication doesn't align with geopolitical context

Input Layers:

Layer	Description	Type
technical	Technical sophistication indicators	float
threat_intel	Threat intelligence confidence	float
operational_impact	Severity of operational impact	float
geopolitical	Geopolitical context alignment	float

Output:

{
  "alert": "string",
  "confidence": "string - high/medium/low",
  "mismatch_explanation": "string",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

4. climate_maladaptation

Name: Climate: Maladaptation Preventer
Description: Blocks solutions that solve immediate micro problems but create macro/meta harms

Input Layers:

Layer	Description	Type
atmospheric	Atmospheric condition metrics	float
ecological	Ecosystem health indicators	float
infrastructure	Infrastructure resilience	float
policy	Policy coherence score	float

Output:

{
  "alert": "string",
  "decision": "string - proceed/caution/block",
  "alternative_suggestion": "string",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

5. legal_precedent_drift

Name: Legal: Precedent Drift Alert
Description: Warns when judicial philosophy shifts threaten current precedent-based strategies

Input Layers:

Layer	Description	Type	Range
black_letter	Statutory text alignment	float	0-1
precedent	Precedent consistency	float	0-1
operational	Practical implementation	float	0-1
socio_political	Social/political context	float	[-1, 1]

Output:

{
  "alert": "string",
  "drift_direction": "string - left/right/fragmenting",
  "strategy_pivot": "string - recommended strategy adjustment",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

6. military_friction_forecast

Name: Military: Friction Forecast Engine
Description: Identifies where tactical plans hit logistics or political constraints

Input Layers:

Layer	Description	Type
maneuver	Tactical maneuver feasibility	float
intelligence	Intelligence confidence	float
sustainment	Logistics sustainability	float
political	Political authorization level	float

Output:

{
  "alert": "string",
  "bottleneck": "string - which layer is the constraint",
  "risk_rating": "string - high/medium/low",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

7. social_narrative_rupture

Name: Social/Cultural: Narrative Rupture Detector
Description: Catches virality that outpaces institutional sense-making capacity

Input Layers:

Layer	Description	Type	Range
individual	Individual sentiment velocity	float	0-1
network	Network propagation speed	float	0-1
institutional	Institutional response lag	float	0-1
cultural	Cultural archetype alignment	float	[-1, 1]

Output:

{
  "alert": "string",
  "rupture_timing": "string - imminent/ongoing/contained",
  "recommended_adjustment": "string",
  "m_score": "float",
  "spatial_component": "float",
  "layer_attribution": "object"
}

---

Emergence Tools (Confluence Detection)

1. healthcare_precision_therapeutic

Name: Healthcare: Precision Therapeutic Window
Description: Identifies rare alignment of genomic predisposition, environmental readiness, phenotypic timing, and psychosocial engagement for maximum treatment efficacy

Input Layers:

Layer	Description	Type
genomic_predisposition	Genetic readiness for treatment (0-1)	float
environmental_readiness	Exposure/toxin levels favorable (0-1)	float
phenotypic_timing	Disease progression stage optimal (0-1)	float
psychosocial_engagement	Patient motivation/support high (0-1)	float

Output:

{
  "window_detected": "boolean",
  "window_type": "string - OPTIMAL/FAVORABLE",
  "confidence": "float 0-1",
  "m_score": "float",
  "alignment_floor": "float - minimum layer value",
  "limiting_factor": "string - weakest layer",
  "recommended_action": "string",
  "duration_estimate": "string"
}

Confluence Threshold: 0.65 (all layers must exceed)

---

2. finance_confluence_alpha

Name: Finance: Confluence Alpha Engine
Description: Detects asymmetric opportunity when technical setup, macro tailwind, flow positioning, and risk compression achieve directional harmony

Input Layers:

Layer	Description	Type	Range
technical_setup	Technical indicators favorable	float	0-1
macro_tailwind	Fundamental/macro support	float	0-1
flow_positioning	Crowd positioning (extreme = signal)	float	[-1, 1]
risk_compression	Risk appetite favorable	float	0-1

Output:

{
  "window_detected": "boolean",
  "setup_quality": "string - HIGH/MODERATE_CONVICTION",
  "conviction_score": "float 0-1",
  "edge_source": "string",
  "recommended_action": "string",
  "stop_loss": "string",
  "m_score": "float"
}

---

3. cyber_adversary_overreach

Name: Cybersecurity: Adversary Overreach Detector
Description: Identifies defensive advantage windows when attacker TTPs are stretched, geopolitically pressured, and operationally fatigued

Input Layers:

Layer	Description	Type
threat_intel_stretch	Attacker TTPs overextended/visible (0-1)	float
geopolitical_pressure	External pressure on attacker (0-1)	float
operational_hardening	Defender readiness/hardening (0-1)	float
tool_reuse_fatigue	Attacker tool reuse/indicators (0-1)	float

Output:

{
  "window_detected": "boolean",
  "attacker_state": "string - RESILIENT/STRESSED/OVEREXTENDED",
  "defender_advantage": "string - LOW/MODERATE/HIGH/CRITICAL",
  "attacker_strain_score": "float 0-1",
  "recommended_action": "string",
  "duration_estimate": "string",
  "counter_attack_viability": "string",
  "m_score": "float"
}

---

4. climate_resilience_multiplier

Name: Climate: Resilience Multiplier
Description: Surfaces interventions with positive cross-domain coupling solving multiple layer problems simultaneously

Input Layers:

Layer	Description	Type
atmospheric_benefit	Atmospheric/climate benefit (0-1)	float
ecological_benefit	Ecosystem benefit (0-1)	float
infrastructure_benefit	Infrastructure resilience benefit (0-1)	float
policy_alignment	Policy coherence/support (0-1)	float

Output:

{
  "window_detected": "boolean",
  "intervention_type": "string - MULTIPLIER/HIGH_MULTIPLIER",
  "cross_domain_coupling": "float 0-1",
  "benefit_layers_above_70": "integer 0-4",
  "example_intervention": "string",
  "funding_priority": "string",
  "m_score": "float"
}

---

5. legal_precedent_seeding

Name: Legal: Precedent Seeding Optimizer
Description: Spots windows when socio-political climate, institutional capacity, statutory ambiguity, and circuit splits align for favorable case law establishment

Input Layers:

Layer	Description	Type
socio_political_climate	Receptiveness to legal change (0-1)	float
institutional_capacity	Courts/resources to handle case (0-1)	float
statutory_ambiguity	Statutory text ambiguity/openness (0-1)	float
circuit_split	Degree of circuit split (0-1)	float

Output:

{
  "window_detected": "boolean",
  "precedent_opportunity": "string - LOW/MODERATE/HIGH/EXCEPTIONAL",
  "ripeness_score": "float 0-1",
  "circuit_split_exploitable": "boolean",
  "strategy": "string",
  "forum_recommendation": "string",
  "timeline": "string",
  "m_score": "float"
}

---

6. military_strategic_initiative

Name: Military: Strategic Initiative Window
Description: Identifies decisive action opportunities when intelligence ambiguity, positional advantage, logistic readiness, and political authorization synchronize

Input Layers:

Layer	Description	Type
enemy_ambiguity	Intelligence gaps favoring surprise (0-1)	float
positional_advantage	Geographical/tactical position (0-1)	float
logistic_readiness	Sustainment capability ready (0-1)	float
authorization_clarity	Political authority clear (0-1)	float

Output:

{
  "window_detected": "boolean",
  "maneuver_type": "string - DEFENSIVE_POSTURE/TACTICAL_INITIATIVE/OFFENSIVE_OPERATION/DECISIVE_ACTION",
  "initiative_score": "float 0-1",
  "advantage": "string",
  "recommended_action": "string",
  "execution_window": "string",
  "risk_assessment": "string",
  "m_score": "float"
}

---

7. social_catalytic_alignment

Name: Social/Cultural: Catalytic Alignment Detector
Description: Spots transformative potential when individual readiness, network bridges, policy windows, and paradigm momentum converge

Input Layers:

Layer	Description	Type
individual_readiness	Population psychological readiness (0-1)	float
network_bridges	Cross-cutting network connections (0-1)	float
policy_window	Policy opportunity open (0-1)	float
paradigm_momentum	Cultural paradigm shift underway (0-1)	float

Output:

{
  "window_detected": "boolean",
  "movement_potential": "string - LOW/MODERATE/HIGH/TRANSFORMATIVE",
  "catalyst_score": "float 0-1",
  "transformative_potential": "float 0-1",
  "critical_mass_risk": "string",
  "recommended_action": "string",
  "mobilization_urgency": "string",
  "m_score": "float"
}

---

Execution Model

Type: Python function call
Entry Point: mantic_thinking/tools/{suite}/{tool_id}.py::detect
Language: Python 3.8+
Dependencies: numpy

Function Signature

def detect(
    layer1,
    layer2,
    layer3,
    layer4,
    f_time=1.0,
    threshold_override=None,
    temporal_config=None,
    interaction_mode="dynamic",
    interaction_override=None,
    interaction_override_mode="scale",
) -> dict:
    """
    Execute mantic detection for the specific domain.
    
    Args:
        layer1-4: Domain-specific float values (0-1, some support -1 to 1)
        f_time: Temporal kernel multiplier (default 1.0)
        threshold_override: Optional dict of per-tool thresholds (bounded/clamped internally)
        temporal_config: Optional dict for temporal kernel tuning (bounded/clamped, domain-allowed)
        interaction_mode: "dynamic" (default) or "base" interaction coefficients
        interaction_override: Optional per-layer interaction coefficients (list of 4 floats or dict keyed by layer)
        interaction_override_mode: "scale" (default) or "replace"
    
    Returns:
        dict with m_score, spatial_component, layer_attribution, 
        and domain-specific fields
    """

Universal Optional Inputs (All 17 Tools)

All tools accept these optional tuning inputs in addition to their four domain-specific layer inputs:

• threshold_override: dict of threshold-name to float (bounded internally)
• temporal_config: dict configuring temporal kernels (bounded internally; kernel_type must be allowed for the tool domain)
• interaction_mode: "dynamic" or "base"
• interaction_override: either a list of 4 floats (tool layer order) or a dict keyed by layer name; values bounded to [0.1, 2.0]
• interaction_override_mode: "scale" (multiply I_pre elementwise) or "replace" (use override as-is)

Iteration Pattern (Correct)

layer_coupling is computed from layer values (L) only. Changing interaction coefficients (I) will not change coupling.

Use this loop:

• Call tool with defaults (interaction_mode="dynamic", no override)
• Read layer_coupling to identify disagreement/noise between layers (input quality problem)
• If needed:
  • adjust the underlying inputs (preferred), or
  • apply interaction_override to dampen noisy layers / amplify confident layers
• Re-call and compare m_score, layer_attribution, and layer_visibility (these can change with I)

Tool Config Files (YAML)

Each tool has a companion YAML config next to its module with structured guidance:
mantic_thinking/tools/{suite}/{tool_id}.yaml (selection, parameter meaning, interaction tuning examples).

---

Model-Specific Usage

For Claude (Computer Use)

# Read SKILL.md to discover tools
# Import via mantic_thinking/adapters/claude_adapter.py
from mantic_thinking.adapters.claude_adapter import get_claude_tools, execute_tool

tools = get_claude_tools()  # Returns 17 tools in Computer Use format

# Friction tool (risk detection)
result = execute_tool("healthcare_phenotype_genotype", {
    "phenotypic": 0.3, "genomic": 0.9, "environmental": 0.4, "psychosocial": 0.8
})

# Emergence tool (opportunity detection)
result = execute_tool("healthcare_precision_therapeutic", {
    "genomic_predisposition": 0.85, "environmental_readiness": 0.82,
    "phenotypic_timing": 0.88, "psychosocial_engagement": 0.90
})

For Kimi (Native Tools)

# Import via mantic_thinking/adapters/kimi_adapter.py
from mantic_thinking.adapters.kimi_adapter import get_kimi_tools, execute, compare_friction_emergence

tools = get_kimi_tools()  # Returns 17 tools in Kimi native format

# Compare friction vs emergence for same domain
comparison = compare_friction_emergence(
    "healthcare",
    friction_params={"phenotypic": 0.3, "genomic": 0.9, "environmental": 0.4, "psychosocial": 0.8},
    emergence_params={"genomic_predisposition": 0.85, "environmental_readiness": 0.82,
                     "phenotypic_timing": 0.88, "psychosocial_engagement": 0.90}
)
# High M in friction = risk. High M in emergence = opportunity.

For Gemini (Function Declaration)

# Import via mantic_thinking/adapters/gemini_adapter.py
from mantic_thinking.adapters.gemini_adapter import get_gemini_tools, execute_tool

# Get tools in Gemini FunctionDeclaration format
tools = get_gemini_tools()  # Returns [{"function_declarations": [...]}]

# Or get flat list
from mantic_thinking.adapters.gemini_adapter import get_gemini_tools_flat
declarations = get_gemini_tools_flat()

# Execute tool
result = execute_tool("climate_resilience_multiplier", {
    "atmospheric_benefit": 0.75, "ecological_benefit": 0.80,
    "infrastructure_benefit": 0.78, "policy_alignment": 0.82
})

For Codex/OpenAI (Function Calling)

# Import via mantic_thinking/adapters/openai_adapter.py
from mantic_thinking.adapters.openai_adapter import get_openai_tools, execute_tool, get_tools_by_type

# Get all 17 tools
all_tools = get_openai_tools()

# Or filter by type
friction_tools = get_tools_by_type("friction")  # 8 tools
emergence_tools = get_tools_by_type("emergence")  # 8 tools

result = execute_tool("cyber_attribution_resolver", {...})  # Friction
result = execute_tool("cyber_adversary_overreach", {...})   # Emergence

For Ollama (Local/Cloud Models)

# Import via mantic_thinking/adapters/openai_adapter.py (Ollama is OpenAI-compatible)
from mantic_thinking.adapters.openai_adapter import get_openai_tools, execute_tool
import openai

# Point at Ollama's OpenAI-compatible endpoint
client = openai.OpenAI(
    base_url="http://localhost:11434/v1",
    api_key="ollama"  # required but ignored
)

tools = get_openai_tools()

# Works with any Ollama model that supports tools:
# - minimax-m2.1:cloud
# - gpt-oss:20b-cloud  
# - glm-4.7:cloud
# - llama3.1, qwen2.5, etc.

result = execute_tool("social_catalytic_alignment", {
    "individual_readiness": 0.82, "network_bridges": 0.85,
    "policy_window": 0.80, "paradigm_momentum": 0.88
})

---

Input Validation Rules

1. Clamping: All inputs clamped to valid range (0-1 or -1 to 1)
2. NaN Handling: Missing data triggers weight redistribution among available layers
3. Weight Normalization: Weights always sum to 1.0 after any adjustment
4. Deterministic: No randomness - same inputs always produce same outputs

---

Response Schema

All tools return a standardized response:

{
  "m_score": "float - final intensity score",
  "spatial_component": "float - raw S value",
  "layer_attribution": {
    "layer1": "float - percentage contribution",
    "layer2": "float - percentage contribution",
    "layer3": "float - percentage contribution",
    "layer4": "float - percentage contribution"
  },
  "...": "domain-specific fields"
}

Friction Tools Additional Fields

{
  "alert": "string - detection message (null if no alert)"
}

Emergence Tools Additional Fields

{
  "window_detected": "boolean - true if favorable alignment detected",
  "recommended_action": "string - what to do with this window"
}

Layer Visibility (v1.2.0+)

All tools now include layer visibility to aid reasoning:

{
  "layer_visibility": {
    "dominant": "Micro|Meso|Macro|Meta",
    "weights_by_layer": {
      "Micro": "float - aggregated weight",
      "Meso": "float - aggregated weight",
      "Macro": "float - aggregated weight", 
      "Meta": "float - aggregated weight"
    },
    "rationale": "string - why this layer dominates",
    "_note": "Interpretive aid; does not affect M-score"
  }
}

Using Layer Visibility:

• Micro-dominant: Trust immediate signals, check for noise/outliers
• Meso-dominant: Verify local context and environmental factors
• Macro-dominant: Systemic trend; slower but persistent
• Meta-dominant: Long-term adaptation; check baseline drift

Get explanations via adapters:

from mantic_thinking.adapters.kimi_adapter import explain_result

result = execute("healthcare_phenotype_genotype", {...})
explanation = explain_result("healthcare_phenotype_genotype", result)
# Returns reasoning hints based on dominant layer

Layer Coupling (v1.2.3+)

All tools now include layer coupling to reveal agreement and tension between layers:

{
  "layer_coupling": {
    "coherence": 0.43,
    "layers": {
      "technical": {"agreement": 0.5, "tension_with": {"macro": 0.3, "risk": 0.4}},
      "macro": {"agreement": 0.57, "tension_with": {"technical": 0.3}},
      "flow": {"agreement": 0.63},
      "risk": {"agreement": 0.63, "tension_with": {"technical": 0.4}}
    }
  }
}

Fields:

• coherence (0-1): Overall agreement. 1 = all layers agree, 0 = total disagreement.
• layers.<name>.agreement (0-1): How much this layer agrees with all other layers.
• layers.<name>.tension_with: Only present when pairwise agreement < 0.5. Names the conflicting layer and its agreement score.

Using Layer Coupling for Reasoning:

• High coherence on an emergence window = the window is real (all layers confirm)
• Low coherence on a friction alert + tension pairs = identifies exactly where the conflict is
• No tension_with for a layer = that layer agrees with all others

Note: layer_coupling does not affect the M-score. Like layer_visibility, it is a read-only reasoning hint.

---

Interpretation Guide

Same M-Score, Opposite Meaning

M-Score	Friction Tool Meaning	Emergence Tool Meaning
0.1-0.3	Low risk, systems aligned	Low opportunity, wait
0.4-0.6	Moderate friction detected	Favorable window forming
0.7-0.9	High risk, urgent action needed	Optimal window, act now

Key Principle: The M-score measures intensity. Friction tools interpret high intensity as danger. Emergence tools interpret high intensity as opportunity.

Temporal scaling: When f_time > 1.0, m_score can exceed 1.0. The table above assumes f_time = 1.0 (default) and k_n = 1.0. Interpret scores above 1.0 as amplified intensity: the signal is strengthening over time.

---

Testing

Run cross-model validation:

cd mantic-thinking
python -m pytest tests/test_cross_model.py -v

Expected: All 17 tools pass with 3 test cases each.

Quick sanity check:

python -c "from mantic_thinking.adapters.openai_adapter import get_openai_tools; print(len(get_openai_tools()), 'tools available')"