Skip to main content

​
Overview

Templates are the “patches” that define reproducible simulation scenarios in Timepoint Pro. Like presets on a synthesizer, they capture a complete configuration—entities, temporal mode, fidelity settings, and output formats—that can be loaded and run repeatedly.

​
Template Structure

Templates are JSON files stored in generation/templates/ with the following structure: 
{
  "scenario_description": "High-level description of the scenario",
  "world_id": "unique_identifier",
  "patch": {
    "name": "Display Name",
    "category": "corporate|historical|crisis|mystical|etc",
    "tags": ["conflict", "decision", "negotiation"],
    "author": "your-name",
    "version": "1.0",
    "description": "Detailed description of what this template does"
  },
  "entities": { ... },
  "timepoints": { ... },
  "temporal": { ... },
  "outputs": { ... }
}

​
Key Sections

​
Entities Configuration

Defines the population of entities in your simulation: 
"entities": {
  "count": 5,
  "types": ["human"],
  "initial_resolution": "tensor_only",
  "animism_level": 0,
  "envelope": {
    "attack": 0.15,
    "decay": 0.1,
    "sustain": 0.8,
    "release": 0.2
  },
  "adprs_envelopes": [
    {
      "A": 0.7,
      "D": 31536000000,
      "P": 2.0,
      "R": 0,
      "S": 0.8,
      "t0": "2026-01-01T00:00:00+00:00",
      "baseline": 0.1
    }
  ]
}
Key fields:
  • count: Number of entities to generate
  • types: Entity types ("human", "animal", "building", "kami", "abstract", "any", "ai")
  • initial_resolution: Starting fidelity level ("tensor_only", "basic_profile", "full_context")
  • envelope: ADSR envelope for entity lifecycle (Attack, Decay, Sustain, Release)
  • adprs_envelopes: SynthasAIzer waveform envelopes for temporal state evolution

​
Timepoints Configuration

Defines the temporal structure: 
"timepoints": {
  "count": 3,
  "start_time": null,
  "resolution": "hour",
  "before_count": 0,
  "after_count": 0
}
Key fields:
  • count: Number of timepoints to simulate
  • resolution: Time granularity ("second", "minute", "hour", "day", "week")
  • before_count/after_count: Optional timepoints before/after the main sequence

​
Temporal Mode Configuration

Defines causality model and mode-specific parameters: 
"temporal": {
  "mode": "forward",
  "enable_counterfactuals": false,
  "fidelity_planning_mode": "hybrid",
  "token_budget_mode": "soft",
  "token_budget": 80000.0,
  "fidelity_template": "balanced"
}
Temporal modes:
  • "forward": Strict forward causality (default)
  • "portal": Backward reasoning from a target state
  • "branching": Counterfactual timeline splits
  • "cyclical": Seasonal/cyclical patterns with prophecy
  • "directorial": Dramatic tension drives causality
Portal mode specific: 
"temporal": {
  "mode": "portal",
  "portal_description": "Failed Mars mission with life support failure",
  "portal_year": 2031,
  "origin_year": 2026,
  "backward_steps": 15,
  "exploration_mode": "adaptive",
  "coherence_threshold": 0.7,
  "use_simulation_judging": false
}

​
Output Configuration

Defines what artifacts to generate: 
"outputs": {
  "formats": ["json", "markdown"],
  "include_dialogs": true,
  "include_relationships": true,
  "include_knowledge_flow": true,
  "export_ml_dataset": false,
  "generate_narrative_exports": true,
  "narrative_export_formats": ["markdown", "json", "pdf"],
  "narrative_detail_level": "summary",
  "enhance_narrative_with_llm": true
}

​
Example Templates

​
Board Meeting (Forward Mode)

From generation/templates/showcase/board_meeting.json: 
{
  "scenario_description": "Tech startup board meeting where CEO proposes an acquisition",
  "world_id": "board_meeting_example",
  "patch": {
    "name": "Boardroom Brass",
    "category": "corporate",
    "tags": ["conflict", "decision", "multi-entity", "negotiation"]
  },
  "entities": {
    "count": 5,
    "types": ["human"],
    "initial_resolution": "tensor_only"
  },
  "timepoints": {
    "count": 3,
    "resolution": "hour"
  },
  "temporal": {
    "mode": "forward"
  }
}
Mechanisms exercised: M1 (Heterogeneous Fidelity), M7 (Causal Chains), M11 (Dialog Synthesis), M13 (Relationship Evolution) Cost estimate: ~$0.05 | Duration: ~2 minutes

​
Mars Mission Portal (Portal Mode)

From generation/templates/showcase/mars_mission_portal.json: 
{
  "scenario_description": "Portal backward reasoning from failed Mars mission to root causes across 5 years",
  "temporal": {
    "mode": "portal",
    "portal_description": "Ares III Mars mission fails in 2031 with life support failure during EVA",
    "portal_year": 2031,
    "origin_year": 2026,
    "backward_steps": 15,
    "exploration_mode": "adaptive",
    "coherence_threshold": 0.7
  }
}
Mechanisms exercised: M17 (Modal Causality), M3 (Exposure Events), M7 (Causal Chains), M8 (Embodied States), M11 (Dialog), M13 (Relationships) Cost estimate: ~$0.40 | Duration: ~15 minutes

​
Creating a New Template

​
Step 1: Start with a Base

Copy an existing template from generation/templates/showcase/ that matches your use case: 
cp generation/templates/showcase/board_meeting.json \
   generation/templates/showcase/my_scenario.json

​
Step 2: Edit Core Fields

Update the scenario description, world ID, and patch metadata: 
{
  "scenario_description": "Your scenario description",
  "world_id": "my_unique_scenario_id",
  "patch": {
    "name": "Your Scenario Name",
    "category": "your_category",
    "tags": ["relevant", "tags"],
    "author": "your-name",
    "version": "1.0",
    "description": "Detailed description"
  }
}

​
Step 3: Configure Entities

Adjust entity count, types, and resolution: 
"entities": {
  "count": 8,
  "types": ["human", "animal", "building"],
  "initial_resolution": "basic_profile",
  "animism_level": 2
}

​
Step 4: Set Temporal Mode

Choose the appropriate causality model: 
"temporal": {
  "mode": "branching",
  "enable_counterfactuals": true,
  "fidelity_template": "high_detail"
}

​
Step 5: Register in Catalog

Add your template to generation/templates/catalog.json: 
"templates": {
  "showcase/my_scenario": {
    "name": "My Scenario",
    "description": "Brief description",
    "mechanisms": ["M7", "M11", "M13"],
    "tier": "standard",
    "category": "showcase",
    "cost_estimate": "$0.10",
    "duration_estimate": "4min",
    "status": "pending"
  }
}

​
Step 6: Test Your Template

Run your template and verify it works: 
./run.sh run my_scenario

​
Template Categories

Templates are organized by category in the catalog:
  • corporate: Business scenarios (board meetings, VC pitches)
  • historical: Historical events (Jefferson Dinner)
  • crisis: High-stakes emergencies (hospital crisis)
  • mystical: Supernatural/spiritual scenarios (kami shrine)
  • mystery: Detective/investigation scenarios
  • portal: Portal mode backward reasoning
  • space: Space exploration scenarios
  • legal: Legal proceedings and investigations
  • persona: Agent persona demonstrations

​
Template Tiers

​
Quick Tier

  • Entities: 2-3
  • Timepoints: 1-2
  • Cost: Less than $0.05
  • Duration: Under 1 minute
  • Use case: Fast validation tests

​
Standard Tier

  • Entities: 4-6
  • Timepoints: 3-5
  • Cost: 0.05−0.05-0.20
  • Duration: 1-5 minutes
  • Use case: Normal testing with moderate complexity

​
Comprehensive Tier

  • Entities: 7-10
  • Timepoints: 5-10
  • Cost: 0.20−0.20-1.00
  • Duration: 5-15 minutes
  • Use case: Full mechanism exercise with rich scenarios

​
Best Practices

​
Fidelity Management

Start with low resolution and let the system upgrade on demand: 
"entities": {
  "initial_resolution": "tensor_only"  // ~200 tokens per entity
}
Use fidelity templates for common patterns: 
"temporal": {
  "fidelity_template": "balanced",     // Balanced quality/cost
  "token_budget_mode": "soft",         // Allow budget overruns
  "token_budget": 80000.0
}

​
Cost Optimization

  1. Use "initial_resolution": "tensor_only" to minimize token usage
  2. Set realistic token_budget based on your needs
  3. Limit timepoints.count to what you need
  4. Use "export_ml_dataset": false unless generating training data
  5. Set "enhance_narrative_with_llm": false to skip LLM-enhanced narratives

​
Temporal Mode Selection

  • Use forward for standard simulations with strict causality
  • Use portal when you know the outcome and want to trace causes backward
  • Use branching to explore “what if” scenarios at decision points
  • Use cyclical for seasonal patterns or repeating prophecies
  • Use directorial when dramatic tension should drive events

​
Troubleshooting

​
Template Won’t Load

Validate JSON syntax: 
python -m json.tool generation/templates/showcase/my_scenario.json

​
Template Runs Too Long

Reduce complexity:
  • Decrease entities.count
  • Decrease timepoints.count
  • Use "initial_resolution": "tensor_only"
  • Set "token_budget_mode": "hard" with lower budget

​
Template Too Expensive

Optimize for cost:
  • Use "fidelity_template": "minimal"
  • Set "include_dialogs": false if dialog not needed
  • Reduce timepoints.count
  • Set "enhance_narrative_with_llm": false

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us