标签：#harness

A deep dive into what Anthropic, OpenAI, Perplexity and LangChain are actually building. Covering the orchestration loop, tools, memory, context management, and everything else that transforms a stateless LLM into a capable agent. You've built a chatbot. Maybe you've wired up a ReAct loop with a few tools. It works for demos. Then you try to build something production-grade, and the wheels come off: the model forgets what it did three steps ago, tool calls fail silently, and context windows fill up with garbage. The problem isn't your model. It's everything around your model. LangChain proved this when they changed only the infrastructure wrapping their LLM (same model, same weights) and jumped from outside the top 30 to rank 5 on TerminalBench 2.0. A separate research project hit a 76.4% pass rate by having an LLM optimize the infrastructure itself, surpassing hand-designed systems. That infrastructure has a name now: the agent harness. What Is the Agent Harness? The term was formalized in early 2026, but the concept existed long before. The harness is the complete software infrastructure wrapping an LLM: orchestration loop, tools, memory, context management, state persistence, error handling, and guardrails. Anthropic's Claude Code documentation puts it simply: the SDK is "the agent harness that powers Claude Code." OpenAI's Codex team uses the same framing, explicitly equating the terms "agent" and "harness" to refer to the non-model infrastructure that makes the LLM useful. The canonical formula, from LangChain's Vivek Trivedy: "If you're not the model, you're the harness." Here's the distinction that trips people up. The "agent" is the emergent behavior: the goal-directed, tool-using, self-correcting entity the user interacts with. The harness is the machinery producing that behavior. When someone says "I built an agent," they mean they built a harness and pointed it at a model. Beren Millidge made this analogy precise in his 2023 essay, Scaffolded LLMs as Natu

Harness Base Definition: The Control System Outside the Model Previously, we split Agent into several minimal parts: Model: judge the next step Loop: keep the process moving Tools: interact with the real world State: keep the task connected At this point, a natural question appears: If Agent already has model, loop, tools, and state, why talk about Harness? An even easier confusion is: Is Harness a higher-level, smarter Agent that manages other Agents? That sounds plausible, but it bends the architecture in the wrong direction. Harness is not another Agent. It is not a larger prompt, and it is not a framework name. It is the control system outside the model. Continue with the same small CLI Agent: User says: help me figure out why this project's tests are failing, and fix it. If this CLI Agent is only a demo, it can be simple: send user input to model model says read file program reads file put result back into prompt model says edit file program edits file model says run tests program runs tests This chain can work once and already look like an Agent. But as soon as someone else really uses it, questions appear. What if the model wants to execute rm -rf ? What if it wants to read private files under the user's home directory? If it runs for ten minutes and the user interrupts, how is the working state saved? After a tool error, should the next model turn see the full log or only a summary? If the same task continues tomorrow, where does the session resume from? If a modification looks successful but no test verified it, how does the system know it is done? If a user says the Agent damaged a file, how do we reconstruct what happened? These questions do not belong to the model itself. They should not be left for the model to decide. The model only generates the next-step judgment from the current context. Permission, execution environment, session lifecycle, observability logs, verification criteria, and governance policy are engineering responsibilities outside the

Over the past decade, the core evolution of data engineering has been the deconstruction and reconstruction of traditional data warehouse architectures through the Modern Data Stack. We separated data ingestion from databases, forming the Data Ingestion layer, using tools like FiveTran, Airbyte, and Apache SeaTunnel to solve ELT / CDC / Reverse ETL problems; We separated compute from storage, forming cloud data warehouse and lakehouse systems such as Snowflake, Databricks, Iceberg, and Hive; We separated orchestration from scripts, leading to orchestration systems like Apache Airflow and Apache DolphinScheduler; SQL development, data modeling, lineage, data quality, BI, and AI analytics were further split into independent tools. This architecture was undoubtedly progress. It moved data engineering away from the primitive era of “a bunch of scripts + Crontab” toward cloud-native infrastructure, elastic computing, engineering governance, and open ecosystems. The greatest contribution of the Modern Data Stack was “decoupling,” and its biggest side effect was also “decoupling.” Tools became more powerful, but data engineers were forced to switch between more systems than ever before: datasources in one place, synchronization configs in another, DAGs somewhere else, logs elsewhere, SQL stored in Git, and Snowflake / Iceberg / cloud warehouse execution results living in yet another environment. As a result, many data engineers spend less time on data modeling, business understanding, metric definitions, architecture design, and cost optimization — and far more time configuring datasources, setting field mappings, dragging DAG nodes, modifying SQL, checking logs, and rerunning tasks. This is the hidden pain created by the Modern Data Stack: data engineers became trapped inside tools. The emergence of engineering-focused AI systems like Codex and Claude Code is now changing the entire software engineering workflow. But how can data engineers truly achieve Vibe Coding? That