Trace runs block by block
Studio captures every step of a run: each block, tool call, and model response, with its duration.
Logs is the AI agent observability layer in Studio, the open-source AI workspace where teams build, deploy, and manage AI agents. Studio records every step of every run, so teams trace each decision, tool call, and output, and catch failures with alerts.
Logs
Trace every agent run with full observability in Studio.
Logs is the AI agent observability layer in Studio, the open-source AI workspace. Follow each run block by block, search across every run, and catch failures with alerts.
Companies building AI agents with Studio
See exactly what every AI agent did.
Studio records every step of every agent run, so teams can follow each decision, tool call, and output on one timeline.
Watch runs as they happen
Studio shows each run the moment it starts, with its trigger, duration, and output, so teams see agents working in real time.
Every run on the record
Studio keeps a permanent history of every run and follow-up, so teams can answer what happened, when, and who resolved it.
Catch failures before they spread.
Studio surfaces failed runs the moment they happen, with the alerts and context teams need to fix them fast.
Filter to the runs that matter
Search and filter every run in Studio by workflow, status, trigger, and time, so failed runs surface fast.
Debug with full context
Studio snapshots each run, so teams open a failed run with inputs, outputs, and trace intact and see which block broke.
Catch failures with alerting
When a run fails, Studio alerts the channel your team already watches, naming the failing block and linking straight to the trace.