examples

Operon Examples

This directory contains 108 runnable numbered examples demonstrating the operon_ai library, progressing from basic concepts to complete LLM-powered cell simulations.

Recommended Starting Points

If you are evaluating Operon for practical value rather than formal structure, start here:

• 68_skill_organism_runtime.py if you want the clearest answer to “can this improve a real workflow?” It binds stages to fast vs deep models, preserves structure, and lets you add telemetry without rewriting the skill.
• 67_pattern_first_api.py if you want the shortest path to topology advice, reviewer gates, and specialist swarms without touching the substrate directly.
• 66_epistemic_topology.py if you want to inspect the structural analysis underneath those recommendations.

The older examples still matter, but these two are the fastest way to answer the question most engineers actually care about: does this help me build a better agent system?

For temporal memory scenarios — tracking facts that change over time and reconstructing past belief states:

• 69_bitemporal_memory.py — core valid-time vs record-time divergence with corrections
• 70_bitemporal_compliance_audit.py — multi-fact compliance audit with belief-state reconstruction
• 71_bitemporal_skill_organism.py — multi-stage organism with bi-temporal substrate for auditable shared facts
• 72_pattern_repository.py — register, retrieve, and score reusable collaboration pattern templates
• 73_watcher_component.py — runtime monitoring with signal classification and retry/escalate/halt interventions
• 74_adaptive_assembly.py — task fingerprinting, template selection, automatic organism construction, and outcome recording
• 75_experience_driven_watcher.py — watcher experience pool with intervention history driving future recommendations
• 76_cognitive_modes.py — System A/B cognitive mode annotations with watcher mode balance reporting
• 77_sleep_consolidation.py — sleep consolidation cycle: replay, compress, counterfactual, histone promotion
• 78_social_learning.py — cross-organism template sharing with trust-weighted adoption (epistemic vigilance)
• 79_curiosity_driven_exploration.py — curiosity signals triggering escalation on novel inputs
• 80_developmental_staging.py — developmental stages, capability gating, and critical periods
• 81_critical_periods.py — teacher-learner scaffolding with developmental awareness
• 82_managed_organism.py — one-call managed_organism() wiring the full v0.19-0.23 stack
• 83_cli_stage_handler.py — shell out to external CLI tools (Claude Code, Copilot, ruff) as organism stages
• 84_cli_organism.py — full managed CLI organism from a dict of commands with watcher and substrate
• 85_claude_code_pipeline.py — live 3-stage pipeline (plan → implement → review) using claude --print with context chaining
• 86_swarms_topology_analysis.py — analyze Swarms workflow patterns with Operon's epistemic theorems
• 87_animaworks_role_mapping.py — map AnimaWorks organizational configs to Operon's typed stage system
• 88_deerflow_workflow_analysis.py — analyze DeerFlow 2.0 session configs with Operon's epistemic theorems
• 89_seeded_library.py — seed PatternLibrary from Swarms, DeerFlow, and ACG survey patterns
• 90_hybrid_assembly.py — hybrid assembly with library-first + LLM generator fallback
• 91_deerflow_skill_bridge.py — bidirectional DeerFlow Markdown skill to PatternTemplate conversion
• 92_memory_bridge.py — bridge AnimaWorks and DeerFlow memory into bi-temporal facts
• 93_priming_view.py — multi-channel PrimingView extending SubstrateView
• 94_heartbeat_daemon.py — idle-time consolidation via HeartbeatDaemon
• 95_async_thinking.py — Fork/Join sub-queries within a single stage
• 96_codesign_composition.py — Zardini co-design adapter composition with fixed-point convergence
• 97_ralph_hat_analysis.py — parse Ralph hat configs and analyze event-driven topology with hat-to-stage mapping
• 98_aevolve_workspace_analysis.py — parse A-Evolve workspace manifests, analyze single-agent topology, skills-to-stages
• 99_swarms_deployment.py — compile organism to Swarms SequentialWorkflow config
• 100_deerflow_deployment.py — compile organism to DeerFlow session config
• 101_ralph_deployment.py — compile organism to Ralph hat/event config
• 102_scion_deployment.py — compile organism to Scion grove with containerized agents
• 103_distributed_watcher.py — transport-agnostic watcher signal distribution
• 104_evaluation_harness.py — evaluation harness: 20 tasks × 7 configurations with MockEvaluator
• 105_prompt_optimization_interface.py — PromptOptimizer and EvolutionaryOptimizer protocol interfaces
• 106_workflow_generation_interface.py — WorkflowGenerator and HeuristicGenerator with PatternLibrary registration
• 107_live_evaluation.py — live evaluation with real LLM providers (Gemini API, Claude CLI, Codex CLI)
• 108_meta_evolution.py — meta-evolution of organism configurations with optional LLM proposer

Import Style Guide

All examples should follow this import pattern:

# Standard library imports first
import sys
import time
from dataclasses import dataclass
from pathlib import Path

# Third-party imports
from pydantic import BaseModel

# Operon imports - grouped with parentheses for multiple imports
from operon_ai import (
    ATP_Store,
    Membrane,
    Signal,
    ThreatLevel,
)

# Operon submodule imports - separate import statements
from operon_ai.organelles.nucleus import Nucleus
from operon_ai.providers import MockProvider, ProviderConfig

Example Progression

Range	Focus	Key Concepts
01-07	Basics	Topologies, Budget, Membrane
08-11	Organelles	Mitochondria, Chaperone, Ribosome, Lysosome
12-16	Integration	Complete Cell, Metabolism, Memory, Lifecycle
17-18	Advanced	WAgent, Cell Integrity
19-25	LLM Integration	Real providers, Memory, Tools
26-36	Wiring Diagrams	Visual architecture, Composition
37	Formal Theory	Metabolic Coalgebra, budget-bounded halting conditions
38-41	Healing	Budget Tracking, Chaperone Loop, Regenerative Swarm, Autophagy
42-44	Health & Coordination	Epiplexity, Innate Immunity, Morphogen Gradients
45-47	Practical Applications	Code Review, Codebase Q&A, Cost Attribution
48-55	Orchestration	Multi-motif composition, LLM integration, capstone
56-63	Advanced Biology	Epigenetic coupling, Cell types, Tissue, Plasmids, Morphogens
64-65	Optimization & Providers	Diagram optimization, OpenAI-compatible servers
66	Epistemic Topology	Classification, error bounds, parallelism, recommendations
67	Pattern-First API	Reviewer gates, specialist swarms, topology advice
68	Skill Organisms	Provider-bound stages, fast/deep routing, attachable telemetry
69-71	Temporal Memory	Bi-temporal facts, belief-state reconstruction, compliance audit
72-75	Adaptive Structure	Pattern repository, watcher, adaptive assembly, experience pool
76-79	Cognitive Architecture	Cognitive modes, sleep consolidation, social learning, curiosity
80-81	Developmental Staging	Telomere-driven stages, critical periods, teacher-learner scaffolding
82	Managed Organism	One-call managed_organism() wiring full v0.19–0.23 stack
83-85	CLI Integration	Shell-out stage handlers, CLI organisms, Claude Code pipelines

Running Examples

# Basic examples (no LLM required)
python examples/01_code_review_bot.py

# Practical entry point (no LLM required)
python examples/67_pattern_first_api.py --test

# Provider-bound skill organism (mock providers by default)
python examples/68_skill_organism_runtime.py --test

# LLM examples (requires API key)
ANTHROPIC_API_KEY=sk-... python examples/19_llm_code_assistant.py --demo

Naming Conventions

Method Names

Purpose	Preferred Name	Alternatives (avoid)
Energy consumption	consume()	metabolize, drain
Signal processing	process()	handle, execute
Template creation	create_template()	add_template, register_template
Output validation	fold()	validate, parse

Callback Names

Use on_<event> pattern:

• on_state_change - state transitions
• on_error - error events
• on_complete - completion events

Variable Names

• result - operation outcomes
• response - LLM responses
• signal - input signals
• vitals - health status