# Agents Need Measurement Loops—and a Conductor *By Coding Agents Alpha Tracker • August 5, 2026* Theo’s T3 Code debugging postmortem and Kent C. Dodds’ conductor point to the same practical frontier: agent systems that measure, hand off, and recover instead of merely generating patches. ## 🔥 TOP SIGNAL Theo’s T3 Code performance postmortem is a useful boundary for agentic debugging: a vague Codex request produced a confident diagnosis and a 10,000+ line PR that changed nothing. The breakthrough was to stop asking for a fix and have the agent build a console-driven toggle harness; testing the hypotheses isolated an infinite sidebar opacity animation. The human supplied the hypotheses and validation—the agents were valuable as fast codebase searchers and diagnostic-tool builders, not autonomous diagnosticians. [^1] ## ⚡ TRY THIS - **Turn bug reports into experiments.** Isolate the target in a one-tab browser and use Task Manager: Theo notes that browser tooling is weak for CSS/compositor work, while DevTools changes performance characteristics. Ask the agent to build a console-pasteable toggle for suspected effects, apply all, reset, then flip one feature at a time; separate transitions from animations before editing. Theo got the GPU process down to 3% or less with the toggles applied, back above 25% after reset, and eventually traced the worst offender to the sidebar terminal icon’s pulse. [^1] - **Put a conductor over the fleet.** Kent C. Dodds’ pattern is one supervisor spawning isolated Cursor Cloud Agents, using the Kody Koala MCP for handoffs, creating a PR when a worker stalls, receiving completion messages, and sending a Discord summary at the end. Make each worker’s environment, handoff state, and completion message explicit; the control plane is more valuable than another giant prompt. [^2] - **Use near-free models as sidecars.** Theo says Luna became effectively free after an 80% cost reduction and is using it for T3 Code title generation; he wants it on every prompt for descriptions, feedback, and statuses. Route low-risk metadata and auxiliary outputs there, but measure whether the extra calls improve the workflow before letting cheap inference become unbounded background work. [^3] - **Normalize shared skills with a compatibility shim.** DHH reports that Claude Code still does not natively scan `~/.agents/skills`; his workaround is a symlink, despite the Claude docs acknowledging the pattern. Keep skills in one canonical tree, symlink where needed, and add a fresh-machine discovery check to your agent setup. [^4][^5] ## 📡 WHAT SHIPPED - **LLM 0.32:** Simon Willison’s major CLI/library release sends reasoning traces to stderr (`-R/--hide-reasoning` suppresses them), adds the GPT-5.6 family with GPT-5.6 Luna as the default, and supports server-side Code Interpreter, WebSearch, WebFetch, CodeExecution, and MCP tools. The new `llm openai endpoint` command can run one-off prompts against any OpenAI-compatible endpoint—including a local LM Studio model—without logging them. [^6] It also adds typed `stream_events()` for mixed reasoning/text/tool outputs and tool-chain pause/resume from stored history: primitives worth copying into any coding-agent loop that needs human gates and durable state. [^6] - **OpenWiki 0.3:** A full codebase-wiki prompt rewrite reports a 28.57% relative success increase (35% → 45% at `n=2`), 14% fewer tokens, and 26% fewer tool calls per successful task. Install with `npm install -g openwiki@0.3.0`; treat the numbers as an early, self-reported signal and rerun the eval on your own repositories. [^7] - **Model routers are becoming a coding-agent layer.** Not Diamond Code announced routing across gateways and harnesses, including Claude Code, claiming 20–65% lower cost without a quality hit. Mckay Wrigley sees the larger opportunity in blending “jagged” models into smoother behavior, describes router engineering as a third layer after model and harness engineering, and says DeepSeek V4 Flash was cheap enough to offload roughly a half-dozen tasks from his Fable 5 workflow. The cost claim is vendor-reported; the actionable test is per-task routing and blending, not headline token price. [^8][^9][^10][^11] - **Resilience and fallback primitives:** LangChain says Deep Agents, LangGraph, and LangChain can retry interrupted work, follow a safe recovery path, resume from saved state, and fall back to alternate models. Its Gateway announcement adds fallback rules across models and hosts when a provider fails or rate-limits. This is the right production direction: recovery should be part of the agent runtime, not a human restarting a dead run. [^12][^13] ## 🎬 GO DEEPER - **Video — Theo’s T3 Code performance postmortem.** Focus on the diagnostic-harness segment: the agent becomes useful when the engineer turns competing theories into measurable toggles, then the video explains why infinite compositor animations and layered effects kept the page busy. [^1] [![Fable Broke My App and Couldn't Fix It](https://img.youtube.com/vi/TKlOCjLMNtw/hqdefault.jpg)](https://youtube.com/watch?v=TKlOCjLMNtw&t=714) *Fable Broke My App and Couldn't Fix It (11:54)* - **Repo — [OpenWiki](https://github.com/langchain-ai/openwiki).** Study the prompt rewrite as an example of improving an agent by changing its codebase-understanding instructions rather than swapping models; reproduce the success, token, and tool-call measurements before trusting the tiny `n=2` sample. [^7] - **Agent framework — [llm-coding-agent](https://github.com/simonw/llm-coding-agent).** LLM 0.32’s lower-level work was driven by Datasette Agent and llm-coding-agent; inspect the combination of model/tool mixing, structured streaming, human approval, and resume-from-history rather than treating an agent as a single prompt wrapper. [^6] **Editorial take:** The frontier is shifting from prompt quality to control-plane quality: humans design the measurement, agents build the probes and patches, and supervisors carry state between workers. [^1][^2] --- ### Sources [^1]: [Fable Broke My App and Couldn't Fix It](https://www.youtube.com/watch?v=TKlOCjLMNtw) [^2]: [𝕏 post by @kentcdodds](https://x.com/kentcdodds/status/2084800662857277934) [^3]: [𝕏 post by @theo](https://x.com/theo/status/2084748639470272972) [^4]: [𝕏 post by @dhh](https://x.com/dhh/status/2084623526820970543) [^5]: [𝕏 post by @dhh](https://x.com/dhh/status/2084674748823310721) [^6]: [New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging](https://simonwillison.net/2026/Aug/4/new-release-of-llm) [^7]: [𝕏 post by @BraceSproul](https://x.com/BraceSproul/status/2084665878554243275) [^8]: [𝕏 post by @tomas_hk](https://x.com/tomas_hk/status/2084669945150062619) [^9]: [𝕏 post by @mckaywrigley](https://x.com/mckaywrigley/status/2084681158697337182) [^10]: [𝕏 post by @mckaywrigley](https://x.com/mckaywrigley/status/2084682154240557205) [^11]: [𝕏 post by @mckaywrigley](https://x.com/mckaywrigley/status/2084683536452497805) [^12]: [𝕏 post by @LangChain](https://x.com/LangChain/status/2084626606853685572) [^13]: [𝕏 post by @LangChain](https://x.com/LangChain/status/2084660731211833541)