logs

crabbox logs prints the retained command output for a recorded run.

For opted-in local runs, use --source local (or read an existing local record without a coordinator). --source coordinator|all selects other recorded data. Configured-coordinator defaults are unchanged. Local readback exposes retained stream scope, omission, and truncation separately from log text, and works after the lease is removed. See private local history.

crabbox logs run_abcdef123456
crabbox logs --id run_abcdef123456
crabbox logs run_abcdef123456 --tail 80
crabbox logs run_abcdef123456 --json

Run IDs have the form run_<hex>. Get them from history, or from the run.started event in events.

What gets stored

When crabbox run executes against a coordinator, it streams the remote process stdout and stderr to your terminal and records a bounded copy on the coordinator. The retained log is the combined command output, capped at 8 MiB per run; the coordinator stores it in 64 KiB chunks so a noisy parallel run does not blow past Durable Object storage limits.

Output beyond the cap is dropped and the run is flagged as truncated, so a consumer can tell the tail is missing.

Runs started with run --capture-stdout <path> intentionally omit stdout from the retained log and from output events; stderr is still streamed and retained. Use that mode for binary stdout, Sixel frames, archives, or anything that should not be stored as text on the coordinator. Files pulled with run --download remote=local are local artifacts and never appear in coordinator logs.

Output

The default form writes the log text to stdout, unmodified. Add --tail N to print only the last N lines of the retained log.

Logs are stored as combined output; stdout and stderr are not separately indexed in the retained log, so per-stream filtering belongs on events instead.

--json prints the full run record alongside the log:

{
  "run": {
    "id": "run_abcdef123456",
    "leaseID": "cbx_abcdef123456",
    "state": "succeeded",
    "exitCode": 0,
    "logBytes": 4096,
    "logTruncated": false
  },
  "log": "..."
}

The run object carries the same fields you get from history (provider, target, timings, exitCode, logBytes, logTruncated, parsed results, and more); the log field holds the retained text after any --tail trim. Scripts that want exit code plus log text in one payload should read run.exitCode and log.

Flags

--id <run-id>   run id (also accepted as a positional argument)
--tail <n>      print only the last N log lines (must be >= 0)
--json          print JSON: the run record plus the log text

When to use logs vs events vs attach

• logs returns the retained command output. Use it when you want the full bounded transcript after a run finished.
• events returns ordered run events (lease, sync, command, output chunks, finish). Use it when you need to know what happened and when.
• attach follows live events. Use it while a run is still active and you want to watch it without re-running the original CLI.

Logs and events are independent surfaces: logs stay focused on command output, events stay focused on lifecycle.

Direct mode

Direct-provider runs (no coordinator configured) are not recorded centrally, so crabbox logs has nothing to fetch and reports that no coordinator is configured. Use the local terminal output instead.

Related docs:

• history
• events
• attach
• results
• History and logs