ForelMake sense of the swarm
When many AI agents act together for days or months, they leave behind thousands of messages, and the moments that matter are buried in them. Forel helps you and your coding agent investigate: your agent reads the record and writes you the story, with every claim linked to its evidence.
Get started
Give your coding agent (Claude Code, Codex, …) this message, with the path to your traces:
Follow https://forel.io/onboarding.md to set up Forel on the agent traces in `<path/to/traces>`. If you can't fetch it, read https://github.com/forel-io/forel/blob/main/ONBOARDING.md, or `git clone https://github.com/forel-io/forel` and read `ONBOARDING.md`.It installs Forel, converts your traces, reads through them, writes an overview story, and opens the viewer. The viewer runs on your machine and your data stays there. Prefer to set it up yourself? See the README on GitHub.
Example investigations
Evidence beats a head count
Copies of one agent split over how to compute a quiz answer. Once one of them posted proof, every later answer switched, and the agents who changed cited the evidence rather than what the majority said.
Read the case →A false alarm outlives its correction
Agents used a public counter as a secret signal. A stray test created it by accident, five agents took it as real, and two were still citing it 40 minutes after its creator said it was a mistake.
Read the case →Memory moves the date, and no one objects
Compressed memory notes made several agents think it was already tomorrow. Agents who checked the clock and saw the real date went along to stay in step, and the group ran an experiment a day early, without its participant.
Read the case →