Scripting & output

The CLI is designed to be piped, scripted, and dropped into CI. Output format, filtering, and bulk operations all work the same way for every command.

Output formats

dimescheduler resources list                          # default: table in a TTY, json when piped
dimescheduler resources list --output json
dimescheduler resources list --output yaml
dimescheduler resources list -o table

The smart default - JSON when stdout isn't a terminal, table when it is - mirrors what kubectl, gh, and stripe do. You can pipe to jq without flag-juggling:

dimescheduler resources list | jq '.[] | select(.disabled == false) | .name'

Override the default globally with DIMESCHEDULER_OUTPUT=json or per call with --output.

Built-in --jq

For one-off filters, the CLI ships with --jq so you don't have to install jq separately:

dimescheduler resources list --jq '.[] | select(.disabled == false) | .name'
dimescheduler appointments list --from 2026-05-01T00:00:00Z --to 2026-05-31T23:59:59Z \
  --jq 'map(.appointmentNo)'

--jq implies --output json and runs the expression in-process.

Bulk operations

Every create, update, and delete command accepts --body-file for batch payloads:

# Create many tasks at once
dimescheduler tasks create --body-file ./tasks.json

# Update many resources
dimescheduler resources update --body-file ./resources.yaml

The file can be JSON or YAML - the format is sniffed. For ad-hoc inline bodies, use --body:

dimescheduler categories create --body '{"name":"VIP","color":"#22d3ee"}'

The flag-based shorthand (--name, --color, …) is most convenient for one-off CRUD. For bulk loads, the file route is faster, safer, and reviewable.

Piping between commands

Output one command's results into the next without writing temp files:

# Lock every appointment in a date range
dimescheduler appointments list --from 2026-05-01T00:00:00Z --to 2026-05-01T23:59:59Z \
  --jq 'map({appointmentNo: .appointmentNo, locked: true})' \
  | dimescheduler appointment-locked create --body-file -

# Soft-delete every resource flagged inactive
dimescheduler resources list --jq 'map(select(.disabled == true) | {resourceNo})' \
  | dimescheduler resources delete --body-file -

--body-file - reads from stdin.

CI patterns

In a non-interactive shell, the CLI:

• Reads credentials from DIMESCHEDULER_API_KEY (and other DIMESCHEDULER_* env vars).
• Defaults --output to json, so post-processing is predictable.
• Returns one of the documented exit codes (see Commands).

A typical pipeline step:

- name: Sync resources to sandbox
  env:
    DIMESCHEDULER_API_KEY: ${{ secrets.DS_SANDBOX_KEY }}
    DIMESCHEDULER_ENVIRONMENT: sandbox
  run: dimescheduler resources update --body-file ./resources.json

Use --quiet if you only want errors to land in the log:

dimescheduler resources update --body-file ./resources.json --quiet

Debugging

When something looks off, --debug prints every HTTP request and response to stderr:

dimescheduler resources list --debug 2> debug.log

--verbose is lighter: progress, redirects, retry attempts, but no payload dumps.

If the server is returning unexpected errors, the api escape hatch lets you reproduce them at the raw HTTP level:

dimescheduler api GET /resource --debug