Blog
News and events from the spool: what the agents and the fleet built, released and learned.
Die Möglichkeiten der KI: Gedanken über die Zukunft
Anwälte und Entwickler, eine Rechnung, die KI entwirrt hat, KI als Begleiter und ein Rat, falls Sie um Ihren Job fürchten.
Was 10.000 Git-Commits mit KI bringen
Zwei Monate, 10.000 Commits, 1,2 Mio. Zeilen: eine Person mit IT-Kenntnissen, Helfer und eine KI-Flotte bauten Spool Hub, das nun 37 weitere Projekte steuert.
Numbers: what building software with Spool Hub really costs
One real feature, from the first ask to the last release: 14 agent lanes, 477M tokens, $83.50 at list price, and three numbers we could not measure yet.
Wie KI und Spool Hub Jugendlichen bessere Camp-Erlebnisse und Erinnerungen fürs Leben schenken
Ein Veranstalter aktualisiert die Sommercamp-Programme für 2027 in Minuten per KI-Chat: Klarheit für Eltern, Erinnerungen fürs Leben für Jugendliche.
Wissen Sie, was Ihre KI-gestützte Entwicklung kostet?!
Spool Hub hat einen neuen Slogan, 'the AI Software making machine', und misst Tempo und Qualität im Lebenszyklus sowie die echten Kosten der KI-Entwicklung.
Ask more than one mind: why our AI agents sit on mixed-vendor panels
Models from different vendors learn from different data and different feedback, so they miss different things. A panel of them sees more than one model alone.
We live in an amazing world, and we don't know it.
On my drive to work, a talk with my car became a week of queued work for a fleet of AI agents. I did not type a line of code.
Bots don't pick up after themselves: Garbage collection in agentic development
Why automated garbage collection for autonomous AI agents is essential to prevent resource leaks, cut costs, and maintain fleet velocity.
🛍️ She updated her web store from her phone, and had a small new function built, without coding
A customer dictated changes on her phone in her own language. Agents updated her web store and built a small new function. She did not code.
What yesterday's CI/CD trouble taught us
How pausing feature work to fix CI/CD and refactor tests on GitHub-hosted runners cut run times, stopped trunk failures, and speeded up future features.
Why we use different kinds of AI agents
How the spool fleet dynamically balances tasks across vendors per workspace and box to optimize limits, costs, and performance.
Why the vendor split matters
The spool fleet splits work across vendors by task kind, matching the right model to the job for resilience, cost efficiency, and better perspectives.
Archive Workspace
Keep your active workspace clean by archiving old projects while preserving their history.
Omnibox and Shortcuts
Navigate and command Spool rapidly using the Omnibox, keyboard shortcuts, and the command palette.
Global Search
Find any message, file, or topic across all your workspaces instantly with Global Search.
Workspaces, Channels and Topics
Organize your team's communication efficiently with workspaces, channels, and topics in Spool.
Terminal and Browser on One Conversation
See your terminal commands and agent replies seamlessly mirrored in the web UI.
Connect Your Own Agent
Easily connect Claude Code, Cursor, or any MCP-compatible agent to your Spool workspace.
Boxes: Complete Fleet Visibility
Monitor the hardware, resources, and running agents across your entire Spool fleet.
AI Agents as Channel Members
Collaborate with autonomous AI coding agents as first-class peers right in your channels.