Tag
LangChain announces upcoming LangSmith Roadshow events in Dallas, Boston, and Los Angeles, focusing on agent development and workshops with LangSmith Engine.
The author presents Agro, an open-source cross-platform tool for running AI models locally on mobile and desktop devices, and discusses challenges with hardware constraints for on-device agents while seeking community feedback.
The article discusses the disproportionate impact of interfaces in AI and introduces Headlong, an open source microharness for persistent agents that think continuously.
The author built LocalCode Agent Lab to understand coding agents by implementing components like models, tools, and safety checks from scratch.
Seed is a minimal, self-modifying agent harness that provides the smallest starting point for an AI agent, allowing it to grow its own tools and memory through sessions.
LangChain is hosting a Building Agents with Agents Roadshow in Sunnyvale on August 27th, featuring keynotes, hands-on workshops, and networking opportunities for AI agent developers.
This article presents five design patterns for building long-horizon agent harnesses, with an open-source GitHub repository for implementation.
Highlights an open-source sandbox that lets AI agents suspend and resume quickly, compatible with ADK, LangChain, Claude Code, Codex, and other frameworks.
LangChain Academy released a new free course on autonomous agent improvement using LangSmith Engine, covering the agent development lifecycle from identifying issues to monitoring regressions.
A new workflow allows developers to perform the full agent development lifecycle (build, test, deploy, evaluate, govern) directly from within their coding agent, using Agents CLI and Agent Development Kit.
LangChain announces a summer AMA series on the agent development lifecycle using LangSmith, covering building, deploying, monitoring, improving, and governing agents.
The Harness Handbook is a behavior-centric representation synthesized from agent harness codebases using static program analysis and LLM assistance, helping developers and coding agents locate code implementing specific behaviors. It introduces Behavior-Guided Progressive Disclosure (BGPD) to guide agents from high-level descriptions to relevant implementation details, improving localization accuracy and edit-plan quality.
The author presents a pattern for sharing AI agents by compiling them into sealed WASM modules that only import a single inference function, allowing others to run the agent using their own local model without API keys or security risks.
A tweet thread discussing the iterative process of building evaluations and environments for AI agents, emphasizing human-agent collaboration and the importance of data and verifier design.
Superpowers 6 is released, using Fable 5 to run 25 autonomous experiments, improving build speed by 50% and reducing token costs by 60%, with detailed records of the experimental process and lessons from failures.
The University of Hong Kong Data Intelligence Lab has open-sourced the lightweight agent framework OpenHarness and its built-in agent Ohmo, providing core features like tool calling, memory, multi-agent coordination, and support for platforms like Feishu and Slack.
A user shares positive impressions of the GLM-5.2 model, noting its impressive performance and cost-efficiency compared to Deepseek, and reflects on the educational value and limitations of AI agents for coding and learning.
Discusses the common problem of AI agents performing well in development but failing with real user phrasing, asking how developers bridge this gap.
At the Generative AI Developer Conference, ihower shared about the Harness + Loop Engineering series of articles and slides for Agent development.
The author shares five patterns that consistently kill AI agents: too many jobs per agent, no human-in-the-loop for destructive actions, unstructured outputs, no spend caps, and lack of uncertainty escalation paths. Practical guardrails and a checklist for reliable agent deployment are provided.