Build your first report
A useful result. Clear boundaries. Evidence you can inspect.
This full-text reading copy does not save answers. Open the interactive lesson for saved browser notes.
Build a report you can explain
Turn six synthetic activity entries into an HTML report. Keep the input and tests unchanged.
Open the sample report. Look for 6 entries and 270 minutes, with 180 minutes for Orion and 90 for Lumen.
You will preview the work, approve its scope, run the local build and inspect the result. The practice worker supplies reference code. The local reviewer exercises the review protocol. The tests check the resulting program, and the practice makes no model calls.
For a project that needs more preparation, compare the goal-driven and spec-driven approaches.
Start in an isolated practice folder
Use the terminal and Python environment from course setup.
Complete course setup and keep the extracted folder open in the same terminal, with COURSE_WORK set.
Run the practice setup command below. It creates the sample project in $PRACTICE/project and writes the draft work request to $PRACTICE/task.work.json. Keep this terminal open for the rest of the lesson.
Use the command block for your shell. In the interactive lesson, select Windows PowerShell 5.1 or 7 or macOS/Linux — Bash or Zsh in Command shell. Use the same shell throughout. The course uses its Python executable directly; you do not need to activate the environment.
Look for accepted: false and build_started: false in the preparation result. If preparation fails, stop here. Read the named error and keep the printed diagnostic folder and existing practice folder. If you are unsure how to continue, ask your assistant or teammate to review the printed error and failed step. Agree on the next action before trying another directory.
practice setup
macOS/Linux — Bash or Zsh
PRACTICE="$COURSE_WORK/practice"
"$COURSE_PYTHON" "$COURSE_WORK/practice.py" "$PRACTICE"Windows PowerShell 5.1 or 7
$PRACTICE = Join-Path $COURSE_WORK "practice"
& $COURSE_PYTHON (Join-Path $COURSE_WORK "practice.py") $PRACTICE
Ask what may change and how it will be checked
Preview the request and check the proposed scope.
Run the plan command below. Look for draft, an empty list of missing questions and a checkpoint_digest.
Confirm that the goal and both test criteria match the exercise. The plan may change only report.py. It protects the CSV, expected reports, tests, participant registry and pytest configuration.
If the preview needs a correction, stop and revise the request through the supported workflow. Generate a fresh preview before you accept its scope.
plan preview
macOS/Linux — Bash or Zsh
"$COURSE_PYTHON" -m attune_harness plan --request "$PRACTICE/task.work.json" \
--project "$PRACTICE/project" --config "$PRACTICE/project/participants.json" \
--task-dir "$PRACTICE/task"Windows PowerShell 5.1 or 7
& $COURSE_PYTHON -m attune_harness plan --request "$PRACTICE/task.work.json" --project "$PRACTICE/project" --config "$PRACTICE/project/participants.json" --task-dir "$PRACTICE/task"
Agree to the exact request
Accept the exact draft you reviewed. You authorize execution in the next step.
If the preview matches the exercise, replace the placeholder in PREVIEW_CHECKPOINT with the full checkpoint_digest from your preview. Run the command below.
Look for accepted and a new checkpoint. That records your approval of the reviewed scope. You are ready to decide whether to run the build.
scope acceptance
macOS/Linux — Bash or Zsh
PREVIEW_CHECKPOINT="replace-with-your-full-preview-digest"
"$COURSE_PYTHON" -m attune_harness plan --task-dir "$PRACTICE/task" \
--accept --checkpoint "$PREVIEW_CHECKPOINT"Windows PowerShell 5.1 or 7
$PREVIEW_CHECKPOINT = "replace-with-your-full-preview-digest"
& $COURSE_PYTHON -m attune_harness plan --task-dir "$PRACTICE/task" --accept --checkpoint "$PREVIEW_CHECKPOINT"
What has acceptance established?
Agreement to the displayed scope; execution is still separate. The saved request is accepted. Starting the local command workers is the next decision.
The report has already been implemented and checked. Acceptance records intent, not implementation or test evidence.
Give the local commands a bounded go
Review the local participant commands and authorize the practice build.
Check the two local commands in $PRACTICE/project/participants.json and the effects in the accepted plan. Run build with --allow-external to authorize those commands for this practice run.
Look for completed after the configured checks pass. If the build fails, read the saved error and journal before choosing what to do next.
For real model work, check your selected route in the participant setup and funding guide.
offline build
macOS/Linux — Bash or Zsh
"$COURSE_PYTHON" -m attune_harness build "$PRACTICE/task" --allow-externalWindows PowerShell 5.1 or 7
& $COURSE_PYTHON -m attune_harness build "$PRACTICE/task" --allow-external
Separate a completed task from a trustworthy claim
Check the saved task for the change and test results.
Run status below to inspect $PRACTICE/task. Under Planned tasks, look for report.py in Outputs. Under Host checks, look for Passed beside probe:change and probe:final.
The course compatibility results identify the versions and platforms actually tested, including 77 passing cases per probe in the recorded exercise. Use your own task record to see what ran in your environment.
saved task inspection
macOS/Linux — Bash or Zsh
"$COURSE_PYTHON" -m attune_harness status "$PRACTICE/task" --format markdownWindows PowerShell 5.1 or 7
& $COURSE_PYTHON -m attune_harness status "$PRACTICE/task" --format markdown
Take the checked result into your browser
Generate the report from the unchanged six-row CSV.
Run the report command below. It writes $PRACTICE/activity-report.html beside the project. Look for Saved report to … and open that file in your browser.
Check the count, 270 minutes, project totals and date range against the CSV. Choose a new output filename if you run the command again.
report command
macOS/Linux — Bash or Zsh
"$COURSE_PYTHON" "$PRACTICE/project/report.py" "$PRACTICE/project/data/activity.csv" --format html --output "$PRACTICE/activity-report.html"Windows PowerShell 5.1 or 7
& $COURSE_PYTHON "$PRACTICE/project/report.py" "$PRACTICE/project/data/activity.csv" --format html --output "$PRACTICE/activity-report.html"
Leave the work ready to inspect tomorrow
Explain one result, one boundary and the next decision.
Your result and next decision
Write your answer in your own notes, or use the interactive lesson.
This lesson saves your note in this browser when storage is available. Keep the paths to $PRACTICE/task and $PRACTICE/activity-report.html for the next lessons.
Next: Open Come back tomorrow to return to the saved task.