Add % completion by node charts and flip commit default to no-commit

[jordanpadams]

🗒️ Summary

Adds a normalized "% Completion by Node" section to the burnup chart so that progress across nodes with vastly different product counts can be compared on equal footing. Also changes the script's commit behavior so that running it bare no longer auto-pushes to git — --commit must be passed explicitly.

Changes

• scripts/generate_registry_status_reports.py

  • New % Completion by Node section inserted between Overall Summary and per-node sections in burnup_chart.html, with four charts (all-versions + latest-only × bundles + collections).
  • New makePctChart() JS function builds one Chart.js line per node, normalizing cumulative count / target to 0–100%. Nodes with target=0 are skipped. Dashed grey 100% reference line included.
  • NODE_COLORS palette (12 distinct colors, cycling) for per-node lines.
  • pctCharts object + second update loop in updateAllCharts() so date-range controls apply to pct charts without touching their datasets.
  • TOC updated with "% Completion by Node" anchor.
  • Default changed from --commit to --no-commit; --commit must now be passed explicitly to push.

• CLAUDE.md — updated run instructions to reflect the --no-commit default.

• docs/status/ — regenerated CSVs and HTML with latest registry data.

🤖 AI Assistance

Developed with Claude Code assistance (~90% AI-influenced).

⚙️ Test Data and/or Report

Script run locally against live registry data. burnup_chart.html verified in-browser:

• "% Completion by Node" section renders between Overall Summary and per-node sections
• TOC link scrolls to section
• Date-range controls filter pct charts
• Nodes with zero target excluded

♻️ Related Issues

N/A — standalone improvement to status reporting tooling.

🤓 Reviewer Checklist

• Documentation: CLAUDE.md updated.
• Security: No new external inputs; chart data is embedded Python-generated JSON.
• Testing: Visual/manual verification of chart output.
• Maintenance: No new dependencies.

---

🤖 Generated with Claude Code

[tloubrieu-jpl]

Thanks Jordan, besides the current update, I was wondering if we should have the metric results (.csv or .md) saved in a different branch. Possibly directly in the gh-pages branch as RST. and the CSV staged there as resources.

This is not critical anyway.

[jordanpadams]

@tloubrieu-jpl I was thinking the same, but then it gets a little wonky with roundup, since I don't know how that is generated/deployed? definitely something to investigate.