kaizen
Skill0 starsChanged 7 months agoApache-2.0View on GitHub
Kailash Kaizen - production-ready AI agent framework with signature-based programming, multi-agent coordination, and enterprise features. Use when asking about 'AI agents', 'agent framework', 'BaseAgent', 'multi-agent systems', 'agent coordination', 'signatures', 'agent signatures', 'RAG agents', 'vision agents', 'audio agents', 'multimodal agents', 'agent prompts', 'prompt optimization', 'chain of thought', 'ReAct pattern', 'Planning agent', 'PEV agent', 'Tree-of-Thoughts', 'pipeline patterns', 'supervisor-worker', 'router pattern', 'ensemble pattern', 'blackboard pattern', 'parallel execution', 'agent-to-agent communication', 'A2A protocol', 'streaming agents', 'agent testing', 'agent memory', 'agentic workflows', 'AgentRegistry', 'OrchestrationRuntime', 'distributed agents', 'agent registry', '100+ agents', 'capability discovery', 'fault tolerance', 'health monitoring', 'trust protocol', 'EATP', 'TrustedAgent', 'trust chains', 'secure messaging', 'enterprise trust', 'credential rotation', 'trust verification', 'cross-organization agents', 'agent manifest', 'TOML manifest', 'GovernanceManifest', 'deploy agent', 'FileRegistry', 'introspect agent', 'DAG validation', 'validate_dag', 'schema compatibility', 'cost estimation', 'composition validation', 'catalog server', 'CatalogMCPServer', 'MCP catalog', 'budget tracking', 'BudgetTracker', 'PostureBudgetIntegration', 'posture budget', 'budget threshold', 'L3 autonomy', 'L3 primitives', 'EnvelopeTracker', 'EnvelopeSplitter', 'EnvelopeEnforcer', 'ScopedContext', 'ContextScope', 'ScopeProjection', 'DataClassification', 'MessageRouter', 'MessageChannel', 'DeadLetterStore', 'AgentFactory', 'AgentInstance', 'AgentInstanceRegistry', 'AgentSpec', 'PlanExecutor', 'PlanValidator', 'PlanModification', 'Plan DAG', 'gradient zone', 'agent spawning', 'cascade termination', 'scoped context', 'envelope enforcement', 'plan execution', or 'agent lifecycle'. Also covers L3 integration: 'L3Runtime', 'L3EventBus', 'L3EventType', 'EatpTranslator', 'L3 event system', 'EATP audit events', or 'governance events'. Also covers kaizen-agents governance layer: 'GovernedSupervisor', 'governed agent', 'progressive disclosure', 'governance modules', 'AccountabilityTracker', 'CascadeManager', 'ClearanceEnforcer', 'DerelictionDetector', 'BypassManager', 'VacancyManager', 'kaizen-agents', 'governed multi-agent', 'PACT governance', 'anti-self-modification', 'ReadOnlyView', 'governance security', 'NaN defense', or 'bounded collections'. Also covers kaizen-agents Delegate system: 'Delegate', 'delegate facade', 'typed events', 'TextDelta', 'ToolCallStart', 'DelegateEvent', 'progressive disclosure API', 'run_sync', 'budget tracking', 'multi-provider', 'StreamingChatAdapter', 'adapter registry', 'OpenAI adapter', 'Anthropic adapter', 'Google adapter', 'Ollama adapter', 'tool hydration', 'ToolHydrator', 'search_tools', 'BM25 search', 'incremental streaming', or 'token streaming'.
- Reads credentials
20 other projects use a version of this.
Adopted by
Most recently taken up by
- AtilaMedeiros/skillstheir own version, first seen Nov 16, 2026
- Ivan-propio/claude-skillstheir own version, first seen Nov 16, 2026
- viniciusamelio/skillstheir own version, first seen Nov 8, 2026
- LuisSambrano/antigravity-configtheir own version, first seen Nov 8, 2026
- JuankCadavid/akili-specstheir own version, first seen Nov 2, 2026
- spenserhale/skillstheir own version, first seen Nov 1, 2026
Other versions20 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 20
- JuankCadavid/akili-specs★ 12same name
- kuya-egg/Monozukuri★ 11same name
- xfstudio/skills★ 9same name
- flpbalada/fb-skills★ 8same name
- LuisSambrano/antigravity-config★ 8same name
- hfadhlullah/kaizen★ 6same name
- freskhu/genesis★ 5same name
- JoeCotellese/claude-code-config★ 5same name
- nek1987/auto-agent-harness★ 4same name
- ai4brands-design/claude-skills★ 2same name
10 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository37 other files it gives its agents.
- .claude/rules/agents.mdAGENTS.md
- .claude/guides/claude-code/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- core-sdkSkill
- dataflowSkill
- nexusSkill
- kailash-mcpSkill
- cheatsheetsSkill
- development-guidesSkill
- nodes-referenceSkill
- workflow-patternsSkill
- deployment-gitSkill
- frontend-integrationSkill
- testing-strategiesSkill
- architecture-decisionsSkill
- code-templatesSkill
- enterprise-infrastructureSkill
- validation-patternsSkill
- gold-standardsSkill
- security-patternsSkill
- flutter-patternsSkill
- interactive-widgetsSkill
- enterprise-ai-uxSkill
- conversation-uxSkill
- uiux-design-principlesSkill
- value-auditSkill
- ai-interaction-patternsSkill
- eatp-referenceSkill
- coc-referenceSkill
- 29-pactSkill
- claude-code-patternsSkill
- error-troubleshootingSkill
- kailash-mlSkill
- kailash-alignSkill
- co-referenceSkill
- managementSkill
- projectSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.
0
Common
Most of the 18 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 143
Use AgentRegistry for distributed coordination (100+ agents)
- Very specific: line 144
Never skip signatures, never put structured output in provider_config
“Common” is measured: kept by at least 6 in 10 of the 18 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Kailash Kaizen - AI Agent Framework
Production-ready AI agent framework: signature-based programming, multi-agent coordination, 9 orchestration patterns, 6 autonomy subsystems, distributed coordination, enterprise trust.
Quick Start
from kaizen.core.base_agent import BaseAgent
from kaizen.signatures import Signature, InputField, OutputField
from dataclasses import dataclass
class SummarizeSignature(Signature):
text: str = InputField(description="Text to summarize")
summary: str = OutputField(description="Generated summary")
@dataclass
class SummaryConfig:
llm_provider: str = os.environ.get("LLM_PROVIDER", "openai")
model: str = os.environ["LLM_MODEL"]
temperature: float = 0.7
class SummaryAgent(BaseAgent):
def __init__(self, config: SummaryConfig):
super().__init__(config=config, signature=SummarizeSignature())
result = SummaryAgent(SummaryConfig()).run(text="Long text here...")Pipeline Patterns
from kaizen_agents.patterns.pipeline import Pipeline
pipeline = Pipeline.ensemble(agents=[code, data, writing, research], synthesizer=synth, discovery_mode="a2a", top_k=3)
router = Pipeline.router(agents=[code, data, writing], routing_strategy="semantic")
blackboard = Pipeline.blackboard(agents=[solver, analyzer, optimizer], controller=ctrl, max_iterations=10)9 patterns: Ensemble, Blackboard, Router, Parallel, Sequential, Supervisor-Worker, Handoff, Consensus, Debate.
Skill Files
Quick Start: kaizen-quickstart-template, kaizen-baseagent-quick, kaizen-signatures, kaizen-agent-execution
Agent Patterns: kaizen-agent-patterns, kaizen-chain-of-thought, kaizen-react-pattern, kaizen-rag-agent, kaizen-config-patterns
Multi-Agent & Orchestration: kaizen-multi-agent-setup, kaizen-supervisor-worker, kaizen-a2a-protocol, kaizen-shared-memory, kaizen-agent-registry
Multimodal: kaizen-multimodal-orchestration, kaizen-vision-processing, kaizen-audio-processing, kaizen-multimodal-pitfalls
Advanced: kaizen-control-protocol, kaizen-tool-calling, kaizen-memory-system, kaizen-checkpoint-resume, kaizen-interrupt-mechanism, kaizen-persistent-memory, kaizen-streaming, kaizen-cost-tracking
Observability: kaizen-observability-hooks, kaizen-observability-tracing, kaizen-observability-metrics, kaizen-observability-logging, kaizen-observability-audit
Enterprise Trust (v0.8.0): kaizen-trust-eatp -- Cryptographic trust chains, TrustedAgent, secure messaging, credential rotation, cross-org trust
Agent Manifest & Deploy (v1.3): kaizen-agent-manifest -- TOML manifests, GovernanceManifest, FileRegistry, introspect_agent()
Composition Validation (v1.3): kaizen-composition -- validate_dag(), schema compatibility, cost estimation
MCP Catalog (v1.3): kaizen-catalog-server -- CatalogMCPServer with 11 tools, 14 built-in agents
Budget Tracking (v1.3): kaizen-budget-tracking -- BudgetTracker, PostureBudgetIntegration, threshold callbacks
L3 Autonomy: kaizen-l3-overview, kaizen-l3-envelope, kaizen-l3-context, kaizen-l3-messaging, kaizen-l3-factory, kaizen-l3-plan-dag
Governance: kaizen-agents-governance, kaizen-agents-security
Testing: kaizen-testing-patterns
Key Architecture
Signature-Based Programming: Type-safe interfaces with automatic validation, optimization, and prompt generation.
BaseAgent: Error handling, audit trails, cost tracking, streaming, memory, hooks.
Autonomy (6 Subsystems):
- Hooks -- PRE/POST lifecycle events, 6 builtins, RBAC + Ed25519 + isolation (<0.01ms overhead)
- Checkpoint -- Save/load/fork state, 4 backends (FS/Redis/PG/S3), incremental + compressed
- Interrupt -- 3 sources (USER/SYSTEM/PROGRAMMATIC), graceful vs immediate shutdown
- Memory -- 3-tier: Hot (<1ms, in-memory) / Warm (10-50ms, DB) / Cold (100ms+, S3)
- Planning -- PlanningAgent, PEVAgent, Tree-of-Thoughts
- Meta-Controller -- A2A semantic routing, auto-discovery, fallback strategies
AgentRegistry: O(1) capability discovery, 6 event types, heartbeat health, status management (ACTIVE/UNHEALTHY/DEGRADED/OFFLINE), multi-runtime coordination.
Provider Configuration (v2.5.0)
from kaizen.core.config import BaseAgentConfig
from kaizen.core.structured_output import create_structured_output_config
config = BaseAgentConfig(
llm_provider="openai",
model=os.environ["LLM_MODEL"],
response_format=create_structured_output_config(MySignature(), strict=True),
structured_output_mode="explicit",
)
# Azure
config = BaseAgentConfig(
llm_provider="azure", model=os.environ["LLM_MODEL"],
response_format={"type": "json_object"},
provider_config={"api_version": "2024-10-21"},
structured_output_mode="explicit",
)Azure env vars: AZURE_ENDPOINT, AZURE_API_KEY, AZURE_API_VERSION (legacy names emit DeprecationWarning).
Prompt utils: generate_prompt_from_signature(), json_prompt_suffix() from kaizen.core.prompt_utils.
Integration Patterns
# With DataFlow
class DataAgent(BaseAgent):
def __init__(self, config, db: DataFlow):
self.db = db
super().__init__(config=config, signature=MySignature())
# With Nexus
app = Nexus()
app.register("agent", agent_workflow.build())
app.start()
# With Core SDK
workflow = WorkflowBuilder()
workflow.add_node("KaizenAgent", "agent1", {"agent": my_agent, "input": "..."})Critical Rules
- Define signatures before agents; use
response_format(notprovider_config) for structured output - Set
structured_output_mode="explicit"for new agents - Extend BaseAgent, track costs, enable hooks in production
- Use AgentRegistry for distributed coordination (100+ agents)
- Never skip signatures, never put structured output in
provider_config
Related Skills
01-core-sdk, 02-dataflow, 03-nexus, 05-kailash-mcp, 17-gold-standards
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

