Cognition updated Devin so one session can break down large work and delegate subtasks to worker Devins running in separate VMs. It matters for audits, migrations, and QA runs where one long-context agent is slower than explicit parallelism.
The core change is orchestration. According to Cognition’s launch thread, Devin can now "manage a team of Devins," spawning parallel workers that each run in their own VM instead of forcing one long-running session to do everything serially. The product claim is that Devin will also improve "over time" at breaking work down for a given codebase launch thread.
Cognition’s announcement post adds the implementation detail missing from the tweet: the top-level Devin acts as coordinator, assigning subtasks, monitoring progress, resolving conflicts, and compiling results. The same post says users can message individual workers, schedule follow-ups, pause or terminate sessions, and watch compute usage, which makes this closer to managed multi-agent execution than a simple fan-out job runner announcement.
The launch material is most concrete around workflows that already decompose cleanly. Cognition’s use-case examples calls out parallel QA across web pages, large-scale migrations, security and dependency audits, and codebase refactors. Those are the kinds of tasks where isolated environments matter, because each worker can run commands, tests, and verification independently before handing results back to the coordinating session.
The same day, Cognition said Devin is now available in Poke Poke post. Poke’s linked product page describes a text-first interface for starting Devin sessions, monitoring progress, and collaborating around the work Poke page. That does not add new agent capability by itself, but it does broaden where teams can invoke the new managed-Devins workflow.
CopilotKit shipped hooks that let agents inspect app state and call frontend actions, then paired them with Shadify for ShadCN-based UI composition. It gives embedded agents a cleaner path from chat to in-app behavior.
releaseOpenClaw shipped version 2026.3.22 with ClawHub, OpenShell plus SSH sandboxes, side-question flows, and more search and model options, then followed with a 2026.3.23 patch. Teams get a broader plugin surface, but should patch quickly and review plugin trust boundaries as the ecosystem grows.
releaseCursor shipped Instant Grep, a local regex index built from n-grams, inverted indexes, and Bloom filters that drops large-repo searches from seconds to milliseconds. Faster candidate retrieval shortens the coding-agent loop, especially when ripgrep-style scans become the bottleneck.
breakingChatGPT now saves uploaded and generated files into an account-level Library that can be reused across conversations from the web sidebar or recent-files picker. It removes repetitive re-uploading and makes past PDFs, spreadsheets, and images part of a persistent working context.
breakingEpoch AI says GPT-5.4 Pro elicited a publishable solution to one 2019 conjecture in its FrontierMath Open Problems set, with a formal writeup planned. Treat it as an early milestone worth reproducing, not blanket evidence that frontier models can already automate math research.
Devin can now manage a team of Devins. Devin will break down large tasks and delegate them to parallel Devins that each run in their own VM. Over time, Devin gets better at breaking down and managing tasks for your codebase. Available now for all users.
To learn more and view some sample prompts, check out our announcement here: cognition.ai/blog/devin-can…
You can now use Devin in Poke: poke.com/p/devin
Starting today, personal superintelligence is just one tap away. No download, no signup. Text Poke for free now: Poke.com 🌴 — 0:00 – What's Poke? 0:50 – Introducing Poke Recipes 1:25 – Create a Recipe in 10 seconds 1:43 – Earn on Poke 2:44 – Build with npx
