Welcome

Meet Hud

Hud brings runtime context to agentic rollouts. Every change, whether an agent or a developer wrote it, gets checked against how your code actually runs in production, from PR to fully deployed.

It starts with Hud's runtime code sensor, an SDK you add to each backend service. The sensor runs alongside your code with minimal overhead, counts every call to every function and learns each function's baseline. You don't pick metrics or write thresholds.

On every rollout, Hud uses that runtime context at three points:

• Before the merge, it reviews the PR against how the functions it changes run in production.
• As the new version takes traffic, it compares every function with its baseline and flags any that regress.
• When a function regresses, it already has the evidence, so your coding agent can fix it.

Hud works with whichever coding agent your team uses, like Claude Code, Cursor or Codex.