Claude Code team shows a two-agent workflow for work and rubric-based review
petergyang · x · 2026-07-21
## A two-agent loop for creative review and long-running workflows Thariq from the Claude Code team describes a workflow where one agent does the work and a second agent reviews the output against a rubric. He says this is especially useful for subjective tasks like judging whether a video short is “good,” where a separate verification agent can reduce self-preferential bias. The episode also covers how Claude Code handles long-running loops through `/loop`, `/goal`, and workflow patterns, plus a live demo of editing launch videos with Claude and using HTML artifacts for planning and learning. Thariq says the team cut Claude Code’s system prompt by 80%, arguing that smarter models need less direction, fewer constraints, and fewer examples.
More from coding & agent
- A one-page guide maps AI agents from core concepts to real-world use cases — goyalshaliniuk · 2026-07-21
- OxDeAI opensource protocol moves AI-agent policy checks before execution — docybo · 2026-07-21
- Kimi staff member builds a VR companion with Kimi Code K3 demo — dejavucoder · 2026-07-21
- Kimi K3 took 75 minutes and still failed a simple diagram task, user says — MinusKarma01 · 2026-07-21
- Coding agents need better rules for when to read search summaries or full pages — RhubarbLarge2747 · 2026-07-21
- Chart compares open tickets across Gemini, Codex, Claude, Grok and Opus 3 — nptacek · 2026-07-21