We’re entering an era of agent-first tools: stronger feedback loops between agents, with human judgment where it matters. We keep hearing fr…

We’re entering an era of agent-first tools: stronger feedback loops between agents, with human judgment where it matters. We keep hearing from senior engineers that their mental load is growing, even with AI code review in place. When review happens at the PR stage, humans still carry the work of interpreting findings and driving fixes. We need code quality built into the inner loop: independent & adversarial review while the coding agent is working, so it can investigate and fix issues before opening a PR. Think of the coding agent and the review agent (@QodoAI) as a duo: one generates, the other challenges, each with its own harness and context. They work through findings together. Humans set the swarm setup, the standard, and arbitrate when needed. Let's use human judgment where and when needed. Introducing the first agent-to-agent code review solution: Qodo’s Agentic Toolbox. Get full codebase context, your coding rules, and AI code review, brought directly to your coding agent. Available now for Claude Code, OpenAI’s Codex, and Kiro, with support for any MCP-compatible…

Related

• Mornings feel different lately. Before I even get organized, I already have 5+ agents running. Same thing before I go to sleep. Claude Code,…
• Humans' ALMANAC: A Human Collaboration Dataset of Action-Level Mental Model Annotations for Agent Collaboration
• Confirming Correct, Missing the Rest: LLM Tutoring Agents Struggle Where Feedback Matters Most
• From Agent Loops to Structured Graphs:A Scheduler-Theoretic Framework for LLM Agent Execution
• Calibrating the Evaluator: Does Probability Calibration Mitigate Preference Coupling in LLM Agent Feedback Loops?

Source: Itamar Friedman (X) | 2026-09-09