Audit-ready agent workflows
Framework-specific guides for multi-step agents, RAG, crews, and multi-agent teams. Same sealed decision-time context and hash-chained ledger — pick your stack.
LangChain
Chains and tool agents with one tracer — decision/outcome pairs and public labels for the explorer.
Open LangChain guideLangGraph
Multi-step graphs, routes, and tool calls — seal what each node knew when it decided.
Open LangGraph guideLlamaIndex
Retrieval + generation with proven queries and agent tool steps when you need them.
Open LlamaIndex guideCrewAI
Role-based crews with a single .wrap() — decisions and outcomes chained.
Open CrewAI guideAutoGen
GroupChat and AG2-style teams — capture each agent turn with sealed context.
Open AutoGen guidePrefer a lighter path first? Startups & SME or Personal & Hobby