Piping & Scripting

Every command prints structured JSON to stdout and progress/warnings to stderr. This makes pega safe to use in pipelines:

pega cases get CASE-1 | jq '.id'
pega assignments perform ASSIGN-1 --action Submit --data @form.json > result.json

Interactive mode in pipelines

Even in interactive mode, the final PATCH response is the only thing written to stdout — all prompts, summaries, and progress messages go to stderr. So this works:

pega assignments perform ASSIGN-1 --interactive | jq

Suppressing progress output

Use --quiet to suppress stderr progress messages. Inquirer prompts in interactive mode are not suppressed by --quiet — they're user input, not progress chatter.

pega cases get CASE-1 --quiet | jq '.status'

Exit codes

Scripts can rely on the following exit codes for error handling:

Code	Meaning
0	success (including clean cancellation in interactive mode)
1	API/network error (HTTP 4xx/5xx, timeout, network failure)
2	invalid configuration or arguments
130	interrupted (Ctrl+C in interactive mode)

Example: batch processing

#!/usr/bin/env bash
set -euo pipefail

# Fetch multiple cases and extract their statuses
for case_id in MYAPP-1 MYAPP-2 MYAPP-3; do
  status=$(pega cases get "$case_id" --quiet | jq -r '.status')
  echo "$case_id: $status"
done

The script exits immediately on a non-zero exit code (exit code 1 or 2) thanks to set -e.

CI/CD usage

In continuous integration environments, bypass the token cache entirely to ensure each invocation performs a fresh OAuth exchange:

export PEGA_NO_CACHE=true
export PEGA_BASE_URL=https://your-instance.pega.com
export PEGA_CLIENT_ID=${{ secrets.PEGA_CLIENT_ID }}
export PEGA_CLIENT_SECRET=${{ secrets.PEGA_CLIENT_SECRET }}

pega auth login
pega cases get MYAPP-CASE-1 | jq '.status'

With PEGA_NO_CACHE=true, the CLI never reads or writes ~/.pega-cli/token.json. Every invocation performs a fresh OAuth exchange, preventing stale token issues in CI environments.

The typical pattern is:

1. Set PEGA_NO_CACHE=true to disable caching
2. Export the three required credentials from secrets or environment
3. Run pega auth login to authenticate
4. Chain your commands with && to fail fast on errors