Checkpoint Lab — Command-Line Selection, Tag Expressions, Reruns, and Execution Profiles

1. Checkpoint scenario

You are defining a small release-quality execution interface for the synthetic five-test project from Lesson 2. The team needs three slices: smoke, regression-without-slow, and a qa environment variant. One regression test will fail only when a synthetic flag is enabled. Your job is to prove selection before execution, preserve the failure, perform a controlled failed-only rerun, and document exactly which configuration source wins.

2. Preflight

python --version
python -m robot --version
python -m robot.rebot --version
# Optional only if you choose the RobotCode extension exercise:
robotcode --version

• Required: Robot Framework 7.4.2 stable for the pinned walkthrough; Python 3.8+.
• Optional RobotCode exercise: RobotCode 2.7.0, Python 3.10+.
• No browser, API, database, SSH, Pabot, container, or CI service is required.
• Use only synthetic TARGET_ENV/FORCE_FAILURE values.

3. Build the disposable project

Recreate the two files from Lesson 2 exactly: tests/service.robot with four tests and tests/environment.robot with one test. Before running, write down these predictions:

4. Create a committed core selection contract

# config/smoke.args
--include smoke
--exclude slow
--variable TARGET_ENV:local
--variable EXPECTED_ENV:local
--outputdir results/smoke
tests

State the precedence contract in a README note: “Run from the project root. Argument-file contents are inserted at -A; explicit later CLI options can override applicable earlier single-valued settings. This file contains no secrets.”

5. Slice A — smoke

python -m robot -A config/smoke.args

Verify: exactly two tests selected and both pass. Capture the console summary and results/smoke/output.xml. If count differs, stop and diagnose before continuing.

6. Slice B — regression without slow

python -m robot --include regression --exclude slow \
  --outputdir results/regression-fast tests

Verify: exactly three tests selected: Health Is Smoke, Invoice Is Regression, and Controlled Failure. All pass with default FORCE_FAILURE=no.

7. Slice C — environment variant

python -m robot --include env \
  --variable TARGET_ENV:qa \
  --variable EXPECTED_ENV:qa \
  --outputdir results/env-qa tests

Verify: one PASS. Record separately that env chose the test while CLI variables configured its synthetic state.

8. Inject and repair a selection defect

# Intentionally wrong for this project.
python -m robot --include smokeANDslow --outputdir results/broken-select tests

Predict zero selected tests, observe the diagnostic, and preserve the command. Repair it to the intended semantics—either smokeNOTslow or --include smoke --exclude slow. Do not add --runemptysuite just to silence the failure.

9. Produce the controlled first failure

python -m robot --variable FORCE_FAILURE:yes \
  --outputdir results/initial --output original.xml tests

Prediction: 5 selected, 4 PASS, 1 FAIL. Verify the failing long name and the message. Record the process return code if your shell/workflow exposes it. Do not overwrite this directory.

10. Failed-only rerun

python -m robot --rerunfailed results/initial/original.xml \
  --variable FORCE_FAILURE:no \
  --outputdir results/rerun --output rerun.xml tests

Prediction: exactly one test selected and it passes. Verify its long name matches the original failure. If more than one test runs, stop and inspect the original output/selection identity.

11. Merge without destroying provenance

python -m robot.rebot --merge \
  --outputdir results/final \
  --output merged.xml \
  --log merged-log.html \
  --report merged-report.html \
  results/initial/original.xml results/rerun/rerun.xml

Prediction: merged result contains five tests and final status PASS. Verification must also assert that results/initial/original.xml and results/rerun/rerun.xml still exist. The merged file is derivative evidence.

12. Precedence exercise

Add --outputdir results/from-args to a small argument file and then invoke Robot with a later --outputdir results/from-cli. Predict which one wins based on insertion order, then verify the created directory. Record the resolved command contract. This is a pure filesystem/output test; no external system changes.

# config/precedence.args
--outputdir results/from-args
--test Health Is Smoke

python -m robot -A config/precedence.args \
  --outputdir results/from-cli tests

13. Optional external RobotCode profile exercise

# robot.toml -- optional; do not confuse with core Robot Framework
paths = ["tests"]
output-dir = "results/robotcode"

[profiles.smoke]
extend-includes = ["smoke"]
variables = { TARGET_ENV = "local", EXPECTED_ENV = "local" }

robotcode profiles show smoke
robotcode -p smoke run

If you perform this optional step, your evidence packet must state RobotCode 2.7.0, Python 3.10+, selected profile, resolved output directory, and whether .robot.toml exists. If RobotCode is absent, skip the exercise; the mandatory course objective is already satisfied by core argument files.

14. Evidence packet

15. Verification checklist

• Smoke = 2 PASS.
• Regression without slow = 3 PASS.
• QA environment = 1 PASS.
• Broken smokeANDslow is diagnosed rather than suppressed.
• Original = 5 selected with exactly 1 FAIL.
• Rerun = exactly 1 selected and PASS.
• Merged = 5 final tests and PASS.
• Original/rerun/merged outputs are three distinct retained artifacts.
• Argument-file precedence experiment matches the predicted later override.
• Any RobotCode usage is labeled external and versioned.

16. Cleanup / rollback

Only delete the disposable directory after the evidence packet has been inspected. In a real CI system, retention is governed separately and first-failure artifacts should not be deleted merely because a rerun passed.

cd ..
# PowerShell: Remove-Item -Recurse -Force .\rf-cli-lab
# Bash:       rm -rf ./rf-cli-lab

Knowledge check

17. What Chapter 15 adds to the operating model

Your Robot Framework platform now has an explicit execution contract: source root, selectors, variables, argument/profile provenance, output directories, rerun identity, and merge policy. That is the foundation needed before adding browser state. Chapter 16 will introduce SeleniumLibrary and the Playwright-powered Browser library, where selection still matters but each selected test now owns browser/session/context lifecycle and richer evidence.

References and version anchors

• Robot Framework 7.4.2 User Guide — CLI option syntax, name/tag selection, argument files, reruns, skip/exclude semantics, and result post-processing.
• Robot Framework 7.4.2 on PyPI — stable release and Python compatibility anchor.
• RobotCode configuration and RobotCode CLI reference — optional checkpoint extension only.