Autonomous research workflows with AI agents
Using Claude Code agents to scout literature, compare options and draft research notes, and learning where they help and where they don't.
The problem
A lot of research time goes into finding, reading and comparing things before you can make a decision. Much of that is structured enough to hand to an agent, if you can trust what comes back.
What I built
Workflows that run Claude Code agents on research tasks.
TODO(Sidney): one or two concrete examples (what you asked, how the agents were set up, how you checked their work).
What I learned
TODO(Sidney): where agents saved real time, where they went wrong, and the checks you now build in.