Every used Claude Fable 5.2 to mine hundreds of meetings into a product philosophy doc
every · x · 2026-11-17
The Every team shares hands-on results with Anthropic's newly released Claude Fable 5.2:
- Team member hammermt made a $1,500 video about their evals platform Checks — possible because Fable 5.2 combed through hundreds of meetings and his Slack messages, distilling recurring ideas into a product philosophy compendium.
- Other builds: Kieran Klaassen built Everycut, a video editor with an agent inside; Jack Cheng turned a photo of his toddler's playroom into a racing game.
- Every also ran a 1-hour camp on setting up overnight jobs for the model, knowing when the job is done, and cost trade-offs — Fable 5.2 is slow and token-hungry, so they share when to reach for it vs. faster models.
A real-world sample of delegating long-horizon projects to a frontier model, with cost pragmatics.
More from coding & agent
- Dev calls for an 'omni-feed' product: one filterable pane for all agent actions — DavidWells · 2026-11-17
- macOS 27.3 beta adds granular Full Disk Access controls, making Macs more agent-friendly — film_girl · 2026-11-17
- LangChain adds host allowlists to Managed Deep Agents, blocking injected exfil via proxy — hwchase17 · 2026-11-17
- Goodhart's Law meets AI agents: stop counting PRs, measure the delivery system — rseroter · 2026-11-17
- Running 10 AI agents in parallel: "I'm the bottleneck of my own company" — stewones · 2026-11-17
- Bitter lesson comes for multi-agent systems: RL-trained swarms beat hand-designed orchestration — emax · 2026-11-17