Tag
In an AI Engineer talk, Cognition co-founder Walden Yan recommends against multi-agent systems, saying they are fragile and that the team behind Devin now uses a single agent with coherent context.
Discusses the concept of agent harnesses as compositional generalizers for scaling generalization in models, referencing the RLM harness.
This paper presents cost-effective agent harnesses for ARC-AGI-1 that achieve strong performance using DeepSeek V3.2 without fine-tuning, via an Explorer-Definer Pipeline and a Reflective Orchestrator, achieving 67.25% pass@2 at low cost.
As AI coding models become commoditized, the agent harness—the control loop managing tools and workflows—emerges as the key differentiator. This guide maps the field of harnesses, weighing vendor-native performance against the portability of model-agnostic workflows.
This paper presents ActPlane, a policy engine that enforces safety and effectiveness policies for AI agents at the OS kernel level using eBPF, bridging the semantic gap between natural language policy intent and concrete system actions.
The author shares insights on building effective agent harnesses: the best ones minimize LLM reliance for trivial tasks and reserve LLMs for complex reasoning, distinguishing genuine harnesses from simple wrappers.
A handoff pattern for Claude Code and other AI agent harnesses allows tasks to be delegated to fresh sessions, avoiding usage caps, performance degradation, and high costs by generating a script for another session to execute specific tasks.
A comprehensive breakdown of memory architecture in major AI agent platforms (Claude Code, OpenAI Codex, Copilot, Windsurf, Devin, etc.), discussing how memory is managed, current shortcomings, and future directions.
New research on scaling laws for agent harnesses reveals that most token and tool call volume does not matter; the work introduces an effective approach.
A research paper from PwC finds that grep-style text search, when properly integrated into agent harnesses, can match or beat embedding-based retrieval for coding-agent tasks, suggesting vector databases may not be essential for many use cases.