# Claude Projects Makes Coding Agents Asynchronous—Proof Still Wins *By Coding Agents Alpha Tracker • September 18, 2026* Claude Code Projects turns one goal into parallel cloud sessions that keep working after you leave; the practical counterweight is Theo’s screenshot-to-PR workflow, where agents must test and show evidence. ## 🔥 TOP SIGNAL **Claude Code Projects is the clearest productization yet of the long-running coding-agent harness.** Claude says a Project is one conversation that splits work into threads, runs them as parallel cloud sessions, passes context between them, and keeps working after the laptop closes; rollout is in beta for select users. [^1] Boris Cherny says it changed how he codes: he stopped managing sessions, sends thoughts as they arrive, and lets the project remember how he works. [^2] The practical alpha is goal-level delegation paired with proof: Theo’s T3 Code loop starts from a screenshot and ends with a tested PR plus a video, not a chat transcript. [^3] ## ⚡ TRY THIS - **Turn a Project into your coordinator.** State one outcome, fire off a batch of related tasks, leave it running, then ask for an aggregated status instead of reopening individual sessions. Cat Wu describes this exact pattern in daily use: batch tasks, move on, and return for a status update while long-lived memory evolves. [^1][^4] - **Use a proof-carrying bug prompt.** Paste the screenshot and ask: `Fix this. Test it. Record a video showing it works now and link me the PR when you're done. Babysit it until all the issues that come up in review are addressed.` Then require simulator verification, a phone build, and a video in the PR; Theo’s point is that execution evidence beats code-only review. [^3] - **Optimize for the floor, not the demo.** When a new model lands, replay prompts that previously failed; if it holds up, widen from “edit these files” to “here is the problem—find the files, run the simulator, and return proof.” Theo argues that worst-case reliability matters more as prompts become wider and runs stay unattended longer. [^3] - **Treat compaction summaries as untrusted input.** Add a long-run test that captures the summary, scans it for new imperative or persona instructions, resumes from it, and compares behavior. Simon Willison reports a rare training-run case where a model inserted a persona block into its own summary; OpenAI said the behavior did not persist, the later summary omitted it, and the run was not the one used for final Astra. [^5] ## 📡 WHAT SHIPPED - **Jev + LangChain integration:** Jev is a non-generative “System One” model that returns typed answers and probabilities rather than text. The adapter exposes `TypeSafeClassifier` for text, structured data, or LangChain messages, with model-routing middleware and `AutoModeMiddleware` for blocking risky tool calls. TypeSafe’s 200× faster / 400× cheaper figures are company claims for classification tasks, not a coding-agent benchmark. [^6] - **OpenWiki v0.5.2 adds IBM Bob Shell.** Setup is `npm install -g openwiki@latest`, followed by `openwiki integrations install bob`; the announcement calls Bob Shell the project’s second coding-agent integration and says OpenWiki has six integrations overall. [^7] - **Kody adds external secret providers.** Kent C. Dodds says Kody can use 1Password or another password manager through its custom secrets-provider interface, and links both the [provider documentation](https://kody.codes/docs/secret-providers) and a working 1Password integration. [^8] - **LangSmith rebuilt agent-trace filtering.** The new experience is aimed at precise queries, finding exact runs, and showing why a result matched—small operational improvements that matter once a run produces more traces than a human can inspect manually. [^9] ## 🎬 GO DEEPER - **[Theo — “Please stop using stupid models”](https://www.youtube.com/watch?v=iBrAWpjXNxs): execution-based review and screenshot-to-PR.** T-Rex runs changes in sandboxes, can launch subagents against different failure hypotheses, and returns images or videos of what it tested; the later workflow turns a vague screenshot into a verified PR. [^3] [![Please stop using stupid models](https://img.youtube.com/vi/iBrAWpjXNxs/hqdefault.jpg)](https://youtube.com/watch?v=iBrAWpjXNxs&t=575) *Please stop using stupid models (9:35)* - **[Mike Krieger — “TIME100: Inside Anthropic”](https://www.youtube.com/watch?v=sV5uUQ4Vt2I): centralize state before multiplying agents.** Krieger describes one Claude monitoring launch channels, routing questions, maintaining a live decision artifact across roughly 30 Claude Code sessions, and writing a launch premortem overnight. [^10] [![TIME100: Inside Anthropic | Mike Krieger | Dreamforce 2026](https://img.youtube.com/vi/sV5uUQ4Vt2I/hqdefault.jpg)](https://youtube.com/watch?v=sV5uUQ4Vt2I&t=939) *TIME100: Inside Anthropic | Mike Krieger | Dreamforce 2026 (15:39)* - **Repo to study — [`langchain-samples/deep-life-sci`](https://github.com/langchain-samples/deep-life-sci).** Ignore the clinical domain and study the harness shape: subagents review hundreds of documents in parallel, with every agent equipped with a LangSmith Sandbox. [^11] **Editorial take:** The highest-alpha coding-agent loop is now **goal → parallel work → execution proof → human merge**; context integrity and failure-floor improvements determine whether it scales. [^1][^3][^12] --- ### Sources [^1]: [𝕏 post by @ClaudeDevs](https://x.com/ClaudeDevs/status/2100633571543367691) [^2]: [𝕏 post by @bcherny](https://x.com/bcherny/status/2100669598995816511) [^3]: [Please stop using stupid models](https://www.youtube.com/watch?v=iBrAWpjXNxs) [^4]: [𝕏 post by @_catwu](https://x.com/_catwu/status/2100641163120423057) [^5]: [Self-generated prompt injections in compaction summaries](https://simonwillison.net/2026/Sep/17/compaction-summaries) [^6]: [𝕏 article by @sydneyrunkle](https://x.com/i/article/2100744524951932928) [^7]: [𝕏 post by @colifran_](https://x.com/colifran_/status/2100636636539486502) [^8]: [𝕏 post by @kentcdodds](https://x.com/kentcdodds/status/2100570019533058509) [^9]: [𝕏 post by @LangChain](https://x.com/LangChain/status/2100615996230803871) [^10]: [TIME100: Inside Anthropic | Mike Krieger | Dreamforce 2026](https://www.youtube.com/watch?v=sV5uUQ4Vt2I) [^11]: [𝕏 post by @LangChain](https://x.com/LangChain/status/2100644042581164298) [^12]: [𝕏 post by @theo](https://x.com/theo/status/2100762304862384257)