Alibaba Cloud Open-Sources "AI Agent Handbook," an Enterprise-Grade Full Lifecycle Handbook for Agents
Alibaba Cloud has open-sourced the "AI Agent Handbook," covering five parts—architecture, building, running, governance, and tuning—with Harness at its core to move enterprise Agents from pilot to governable production. The 2026 survey cited in the handbook shows that about 46% of enterprises have completed or are in the process of development, while only about 18% have actually deployed to production; about 90% have context and memory needs, about 63% have multi-model routing and cost governance needs, and fewer than about 8% have trajectory-based automated evaluation. Those paying attention to or implementing MCP total about 37%, and only about 9% have private deployments within the enterprise.
Why read this
AI score 66
The handbook lays out the architecture, governance, and tuning path for enterprise Agents moving from pilot to production, and includes 2026 survey data for reference.
Topics
Body · AI translation
Alibaba Cloud Open-Sources the "AI Agent Handbook": Using Harness to Push Enterprise Agents from Pilot to Governable Production
Enterprise-grade Agent Full Lifecycle Practice Handbook: Architecture, Build, Run, Govern, Tune, plus research and case studies. Open-source address: https://t.co/CzC395KiJq
Why was the handbook rewritten? The preface summarizes three new challenges: turning probabilistic intelligence into reliable productivity capable of handling critical tasks; building it into infrastructure that can be deployed at scale across stability, security, performance, and cost; and getting Agents into core business processes rather than leaving them stranded in isolated pilots. The writing goals also include unifying terminology (Harness, Runtime, Agent, Workflow, etc.) to reduce the confusion caused by "the same word meaning different things to different people" being amplified by models.
2026 Survey Results · Investment is a consensus: completed or in development totals about 46% (about 36% last year); actually deployed to production is about 18%; continuous iteration is even lower. Forms coexist: single Agent about 40%, multi-Agent about 42%, HITL as the primary architecture about 15%—autonomy is a tunable parameter, not a replacement of old by new. Coding scales first: procurement of Coding tools about 57%, general Agent frameworks about 78%, nearly 40% use both; the market is fragmented, and portable engineering abstractions are more stable than betting on a single vendor. Difficulties after building: context/memory needs about 90%; multi-model routing and cost governance about 63%; for evaluation, manual spot-checking about 55%, trajectory-based automated evaluation less than about 8%. Success rate: about one in ten reach 90% or above; about 16% have never quantified their success rate at all. Teams with an evaluation system have a noticeably higher proportion of high success rates than those without (compared in the text). MCP: attention or implementation totals about 37%, private deployment within enterprises only about 9%; what is missing is a registry, identity authentication, version canary releases, auditing, and a gateway—not "whether one understands the protocol." Maturity L1–L4: assisted generation → controlled automation → Agentic Execution → Managed & Optimizing; the second half of the book is basically answering how to move from L2 to L3/L4.
Overall Structure of the Handbook · Architecture (1–2): Agentic Application boundaries, enterprise maturity, reference architecture (components / platform responsibilities / lifecycle) · Build (3–6): with Harness as the core: paradigms and responsibility boundaries; task orchestration and long-horizon advancement; context, state, Memory/Knowledge/Skill; controlled actions (Function Calling, MCP, A2A, HITL, verification loop) · Run (7–12): Runtime and sandboxing; state storage and semantic assets; AI gateway unifying LLM/MCP/Agent traffic; asynchrony and automation; Multi-Agent; distributed communication · Govern (13–16): observability and cost attribution; security (injection, authentication, high-risk authorization, egress); Prompt/Skill/MCP/Agent asset management; pre-launch simulation and quality verification · Tune (17–24): model side (SFT, Agentic RL, distillation, etc.) and Agent-side data flywheel: Trace→Trajectory, golden datasets, Badcase, controlled self-evolution, edge runtime · Practice (25–29 + special topics): R&D efficiency, design engineering, operations and enterprise IT, customer sales operations, GOAI Agent Infra; plus a special topic on the Jev decision model, and a special topic on Personal Agent / Agent as a Service · Outlook (30): from Agentic Application to Agentic OS