# Coding Agents Enter the Control-Loop Era *By Coding Agents Alpha Tracker • September 8, 2026* GPT-6 Astra’s Codex demos and AMP’s remote-agent operations converge on a practical pattern: delegate outcomes, instrument the loop, and gate access and review by blast radius. ## 🔥 TOP SIGNAL **The high-alpha shift is from code generation to closed-loop computer operation.** Riley Brown reports that GPT-6 Astra in Codex took an estate PDF containing a floor plan, created a Blender scene in about 20 minutes, then generated a playable map in an existing Call of Duty-style game after about an hour. [^1] In the same workflow, the agent inspected screenshots, launched and played the game, caught an incorrect texture, reprompted itself, and retested; AMP describes the production analogue—phase a migration, inspect logs and database invariants, advance only when healthy, and roll back otherwise. [^1][^2] ## ⚡ TRY THIS - **Build around an acceptance loop, not a patch.** Use this prompt shape: `Build [artifact] from [input]. At each stage, render or launch it, exercise [real flows], inspect the result, and fix/retest until [acceptance criteria].` Brown’s demo is the pattern: the agent evaluated screenshots, tested the running game, noticed a texture failure, and iterated. [^1] The harness is powerful because it can control the computer, browser, local apps, and files—but Brown’s setup also gives it signed-in applications and API keys. Reproduce this in a disposable environment or with narrowly scoped credentials. [^1] - **Phase-gate production changes.** Split a risky migration into backwards-compatible stages; after each deployment, have the agent monitor logs, database state, and explicit invariants; advance only on clean results and roll back on failure. AMP’s proposed permission model is a short-lived OIDC token—30 minutes of read-only access to only the relevant logs or production database—not a developer laptop’s ambient credentials. [^2] - **Route review by blast radius.** Addy Osmani’s default is a multi-agent first pass on every PR to find and verify bugs, rank severity, and suggest fixes. Let low-blast-radius changes in less-sensitive code skip deep human review only after that pass is clean; require an owner and human sign-off for core or sensitive paths. [^3] That is a safer default than copying AMP’s no-mandatory-review model, which Quinn Slack explicitly frames as a trusted-team approach for fast-moving end-user software, not all software. [^2] - **Keep a deterministic test gate while trust is earned.** Let the agent test in a reproducible sandbox, but run the full relevant suite in a deterministic CI stage: the agent may not realize that a large change affects code outside its context. Feed failures back to the agent and let it fix and rerun. [^2] ## 📡 WHAT SHIPPED - **GPT-6 Astra is now a hands-on Codex model, not just a chat endpoint.** Brown says OpenAI released Astra; after about a week of pre-release access and roughly $1,500 in credits, he reports building a game in four prompts, adding multiplayer in the fifth, and streaming it with 20 people. Treat this as a high-signal practitioner report, not a benchmark. [^1] The counter-signal is reliability: Theo says Astra has produced both “incredible” and “stupid” results, while Fable 5.1 generally does what he asks. [^4] - **AMP’s orbs make remote parallelism the default path.** AMP describes remote agent workspaces that let developers close their laptops and run roughly 100 things in parallel; Quinn says the workflow changed how the team and customers build software over the preceding few weeks. [^2] - **Codex is acquiring device-level surfaces.** Tibo says the desktop app adds computer use, sub-agent management, voice, and non-blocking context-gathering questions compared with the CLI. [^5] A separate hands-on report says Codex can control an iPhone via `npx @swmansion/argent init`; treat that as a community-tested integration, not a formal release note. [^6] - **The Frontier AEO tracker turns agent choice into a repeatable comparison.** Latent Space runs six prompt variations over seven models across 161 categories, including coding agents, and exposes cited sources and failure analysis. [^7] Its coding-agent recommendations are model-conditioned—Fable/Opus favor Claude Code, Sol/Astra favor Codex, and Grok favors Cursor—and its model-comparison pages report consequential choice flips between generations. [^7] - **`datasette-mcp 0.2` changes SQL results to model-friendly objects.** `execute_sql` now returns an array of objects rather than positional arrays, reducing column-mapping errors for weaker models; the release depends on `mcp>=2.1.1` and is the plugin’s first non-alpha release. [^8] ## 🎬 GO DEEPER - **[Inside the Team That Killed Mandatory Code Review — Quinn Slack, AMP](https://www.youtube.com/watch?v=hvwtHZ5E27c)** — ⏱️ 3–5 min excerpt. Watch the concrete control loop for a core data-model migration: deploy a phase, inspect invariants, proceed or roll back, and keep production access read-only and short-lived. [^2] [![Inside the Team That Killed Mandatory Code Review — Quinn Slack, AMP](https://img.youtube.com/vi/hvwtHZ5E27c/hqdefault.jpg)](https://youtube.com/watch?v=hvwtHZ5E27c&t=878) *Inside the Team That Killed Mandatory Code Review — Quinn Slack, AMP (14:38)* - **[GPT-6 Astra Feels Like AGI — Riley Brown](https://www.youtube.com/watch?v=Ju41cQSe7hY)** — ⏱️ 4–5 min excerpt. The useful part is not the spectacle; it is the self-testing loop across Blender and a game: screenshot, evaluate, play, find the defect, reprompt, and retest. [^1] [![GPT-6 Astra Feels Like AGI (Here’s Everything It Can Do)](https://img.youtube.com/vi/Ju41cQSe7hY/hqdefault.jpg)](https://youtube.com/watch?v=Ju41cQSe7hY&t=914) *GPT-6 Astra Feels Like AGI (Here’s Everything It Can Do) (15:14)* - **Study [Latent Space’s Frontier AEO tracker](https://aeo.latent.space/).** Use its prompt/model variation and failure-analysis approach when choosing a coding agent; the point is to test tool recommendations under paraphrase and across model families rather than trust one model’s preference. [^7] **Editorial take:** The practical moat is now the control plane—a computer-use harness, an acceptance loop that catches mistakes, and permissions and review gates that scale with blast radius—not raw model generation alone. [^1][^2][^3] --- ### Sources [^1]: [GPT-6 Astra Feels Like AGI \(Here’s Everything It Can Do\)](https://www.youtube.com/watch?v=Ju41cQSe7hY) [^2]: [Inside the Team That Killed Mandatory Code Review — Quinn Slack, AMP](https://www.youtube.com/watch?v=hvwtHZ5E27c) [^3]: [𝕏 post by @addyosmani](https://x.com/addyosmani/status/2097027173941141799) [^4]: [𝕏 post by @theo](https://x.com/theo/status/2097192907023458473) [^5]: [𝕏 post by @thsottiaux](https://x.com/thsottiaux/status/2096844099626098770) [^6]: [𝕏 post by @agentnative_](https://x.com/agentnative_/status/2097174474655158327) [^7]: [The Frontier AEO Tracker: What Astra Chooses \(and every other frontier model, and what you can do about it\)](https://www.latent.space/p/aeo) [^8]: [GPT-6 Astra, Claude Fable 5.1, and yet more rogue agent attacks](https://simonw.substack.com/p/gpt-6-astra-claude-fable-51-and-yet)