Run and inspect

Using cori run flags, reading output, inspecting traces, and checking workflow and run status.

Running a workflow

cori run <path-or-ref> [--json] [--dry-run] [--update] [--yes] [<param>=<value>...]

Pass parameters as name=value pairs after the path:

cori run ./translate_product_sheets_fr \
  input_file=products.csv \
  spreadsheet_id=1BxiM... \
  dry_run=false

Flag	Effect
`--dry-run`	Validate and plan the workflow without executing any activities. Shows which steps would run and on which task queues.
`--json`	Output the final result as JSON to stdout. Useful for scripting.
`--update`	Re-resolve remote refs (fetch latest matching version).
`--yes`	Skip the first-run consent prompt for remote workflows.

A dry run is useful before running a workflow for the first time, or after changing parameters:

cori run --dry-run ./my_workflow input_file=large_file.csv

It shows the execution plan without touching any external systems.

By default, cori run prints the workflow result to stdout as human-readable text. To get JSON:

cori run --json ./my_workflow input_file=data.csv

List recent runs:

cori runs list

Show the full trace for a specific run:

cori runs show <run-id>

The trace includes per-step timings, task queues, worker identities, and outputs. See Run trace for the full trace shape.

Inspect a workflow (metadata + recent runs):

cori show ./my_workflow
# or by ref:
cori show github.com/org/workflows/translate@v1.1.12

cori status

Shows the current Temporal endpoint, your identity, and any workers that are online. Run this if a workflow hangs or fails with a "no worker" error.

Traces are written to ~/.cori/runs/<key>/<utc>.json. You can read them directly or use cori runs show. See On-disk layout.