clean-core-extensibility-atc-assessment

What this skill does

Parses an SAP ATC findings ZIP export (typically named atc_findings_*.zip) and generates a self-contained HTML dashboard the user can open in a browser or share.

---

How to invoke

The user typically says something like:

• "Analyse this ATC export: path/to/atc_findings_BC1.zip"
• "Generate a clean core dashboard from my ATC results"
• "What's our technical debt score?" (then ask for the ZIP if not provided)

---

Workflow

Step 1 — Locate the ZIP file

If the user has given a path, use it directly. If not, ask:

"Please provide the path to your ATC findings ZIP file (e.g. sources/atc_findings.zip)."

Resolve the path relative to the current working directory or the skill directory.

Step 2 — Generate the dashboard

From the skill directory (skills/clean-core-atc-assessment/), run:

python3 scripts/generate_dashboard.py <path-to-zip> --out <output-path.html>

• --out defaults to atc_dashboard.html. Suggest a descriptive name based on the run title or system (e.g. atc_dashboard_BC1_CW16.html).
• --top is accepted but has no effect on the HTML output (lists are dropdown-driven).

Step 3 — Deliver the result

Tell the user:

• Where the HTML file was saved
• Key headline numbers (total score, potential score, number of objects/packages)
• Any notable observations (e.g. one package dominates the score, high % of successor findings)

Offer to open the file or copy the path for them to open in a browser.

---

Output file

The HTML dashboard is fully self-contained except for Chart.js and SheetJS (loaded from CDN for the bar chart and Excel export). For a fully offline file, replace the CDN <script> tags with local copies — or simply view it while connected.

The file has five tabs:

1. Dashboard — all KPIs, charts, and tables; package and object upgrade summary with clickable rows
2. Pkg Upgrades — upgrade-candidate objects per package (only objects that can improve a level)
3. Obj Upgrades — upgrade-candidate objects by transition type (D→A, D→C, D→B, C→A, C→B, B→A)
4. All Data — full untruncated object, package, and referenced SAP object tables
5. Methodology & Help — explains every calculation with real examples from the data

All dashboard lists are dropdown-driven (adjustable row counts) and the caps auto-scale to the export size so large customers never lose data. Every table with a dropdown also has a ⬇ Excel button that exports the visible rows as a real .xlsx file. See references/DASHBOARD-ENHANCEMENTS.md for the full list of dashboard enhancements.

---

Calculation rules (summary — full detail in the help tab)

KPI	Rule
In-scope checks	CL_YCM_CC_CHECK_API_USAGE, CL_CI_TEST_CRITICAL_STATEMENTS, CL_CI_TEST_ADMISSIBLE_ENHANCM
Technical Debt Score	P1 × 10 + P2 × 5 + P3 × 1 (per object; summed for packages/total)
DPRCTD_SUC excluded from score	Findings with message code DPRCTD_SUC (deprecated with successor) are not counted in P1/P2/P3 — they feed only the Potential Score. All other _SUC codes (SELECT_SUC, TC_CLA_SUC, etc.) are counted normally. Matches RISE Dashboard logic.
Clean Core Level (object)	Worst finding: P1 → Level D, P2 → Level C, P3 → Level B
Objects without findings	Not in export; labelled "not assessed here" — not Level A
Potential Score	Score minus reduction from _SUC findings (including DPRCTD_SUC): Released successor → 0 pts; Classic API or unknown → 1 pt
FUGR/FUNC counting	Findings on a function module sub-object (sub_type=FUNC) are counted under the individual function module (sub_name), not the function group. Other FUGR sub-objects (INCL, etc.) stay under the function group. Matches RISE Dashboard logic.
Composite object key	Objects are identified by (obj_name, obj_type) — same name with different types are counted as distinct objects. Matches RISE Dashboard logic.
Referenced SAP object impact	Sum of weighted finding points across all custom objects referencing it

---

Cloudification Repository data (for potential score)

The successor classification lookup uses bundled JSON files in scripts/data/:

• objectReleaseInfo_PCELatest.json — released SAP objects
• objectClassifications_SAP.json — classic API / noAPI objects
• cr_metadata.json — version record with commit SHA and date

To check for updates (requires internet access, run from laptop):

python3 scripts/update_cr_data.py --check

To download updates:

python3 scripts/update_cr_data.py

See references/HOW-TO-UPDATE-CR-DATA.md for the full guide.

---

Notes

• All score and level values reflect only objects with at least one in-scope ATC finding. Make clear to the user that this is a partial view of the landscape — objects without findings could be Level A but are not confirmed here.
• Out-of-scope findings (CL_CI_TEST_PROGRAM, CL_CI_TEST_CUST_MODIFICATIONS) are counted but excluded from calculations; the dashboard shows the excluded count in the run bar.