Agents & InferenceSimon Willison

llm-keys-ui 0.1

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Match the models (Optional)

Which model wrote which summary? Select a matchup mapping below before voting.

Summary A

A new LLM plugin spins up a local web UI (bound to LAN/Tailscale IPs) so you can enter API keys via browser instead of pasting them into an agent's chat session, keeping secrets out of the transcript. If you're running remote coding agents like Codex, this closes an obvious leak vector—keys stay retrievable via `llm keys get` at runtime without ever appearing in the agent's context or logs.

LinkedIn

Two AI summaries of each story, blind-voted — see today's agents & inference digest →