refactor(pmo-pert): 4-sheet output, PD-only effort, PMI-correct MR#7
Merged
Conversation
Foundation for the 4-sheet workbook refactor. The i18n module ships en/it label dicts and a t(lang, key, **fmt) accessor; config_compat backfills modern fields (pm/devops/alta uplift, calendar) on legacy JSON input and emits a single stderr warning so users know they're on the old format.
Replaces the legacy Resources (%-allocation) and Timeline (sequential Gantt) sheets with a single PD-based role × week matrix. Primary role per leaf is activity.resources[0]; phase weeks come from phase.start_week/end_week or fall back to a sequential layout sized by PERT duration. Cells flag overcommit (>5 PD/week per role) in red and warn (>=4.5 PD) in yellow.
… calendar duration - New rows: Tech PERT, PM Overhead, DevOps Overhead, Subtotal, Contingency, Fascia BASSA, Management Reserve, Fascia MEDIA, Fascia ALTA. MR is computed on the Fascia BASSA cell (Tech+Overhead+Contingency), no longer on Contingency alone (Issue #3, PMI PMBOK §4.3 compliant). - Calendar Duration is a single explicit number derived from config.calendar_total_weeks or phase start_week/end_week (Issue #2). Sequential-sum CI Duration block dropped. - Effort by Team computed from WBS primary role of each leaf in real PD, not as a %-allocation cross-ref to Resources (Issue #1). - Sensitivity Scenarios listed as text rows when provided in config. - Sheet name localized via i18n (Riepilogo in IT).
build() now accepts an optional wbs_info argument. When provided, the
Management Reserve cell uses the PMI PMBOK §4.3-compliant base:
=(WBS!H{total}*(1+pm_pct+devops_pct)+L{contingency_total})*mr_pct
When omitted (legacy callers/tests without WBS), the cell falls back to
the original =L{total}*pct formula. The pipeline (generate_excel.py) is
wired in the next batch to always pass wbs_info.
generate_excel.py now: - routes JSON through helpers.config_compat.normalize_config() (legacy warning + default backfill); - builds WBS, Resource Plan, Risks, Summary in that order; - passes wbs_info into wb_risks.build() so the MR formula picks up the PMI-compliant base. Risks sheet name is now localized (Rischi in IT). Summary cross-references the Risks sheet by its actual name (single-quoted when needed). The validate_template.py script and its tests accept the 4-sheet layout and both English / Italian sheet-name aliases.
SKILL.md, workflow.md, template-criteria.md, excel-schema.md updated to describe the 4-sheet layout (WBS, Resource Plan / Pianificazione Risorse, Risks / Rischi, Summary / Riepilogo) and the v2 JSON schema fields: pm_overhead_pct, devops_overhead_pct, alta_uplift_pct, calendar_total_weeks, phase.start_week/end_week, scenarios[]. Includes a migration example for legacy JSON, the PMI-correct MR formula, the primary-role convention (resources[0]), and Resource Plan capacity rules.
…ation - sample-input.json updated with v2 fields (pm/devops overhead, alta uplift, calendar_total_weeks, phase start_week/end_week, scenarios). - Scenarios 1-5 backfilled with v2 defaults and phase weeks; xlsx regenerated against the 4-sheet generator. - Scenario 6 (PA portal integration) added as an anonymized public- sector scenario in Italian (lang=it). Exercises overlapping phases, 20% MR on the PMI-correct base, PM/DevOps overhead, and the alta uplift band — the new capabilities together. - validation-summary.md rewritten to cover the 4-sheet layout and the five real-world issues that motivated the refactor.
…rate template - Delete scripts/helpers/wb_timeline.py and scripts/helpers/wb_resources.py along with their test modules. Both modules implemented sheets that the v2 refactor removed (Issue #1 — Resources held % presented as effort; Issue #4 — Timeline Gantt was sequential and misleading). - Regenerate assets/pert-template.xlsx by running the v2 generator on the updated sample-input.json so the bundled reference template matches the 4-sheet layout. - Test suite is green (135 tests) and the structural CI test suite passes. - End-to-end smoke: sample workbook opens with 4 sheets, Resource Plan totals are =SUM formulas over per-week PD numerics, Summary shows the Low/Medium/High bands plus Calendar Duration = 12 weeks.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 7d31d29514
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+286
to
+289
| starts = [int(p["start_week"]) for p in phases if "start_week" in p] | ||
| ends = [int(p["end_week"]) for p in phases if "end_week" in p] | ||
| if starts and ends: | ||
| return max(ends) - min(starts) + 1 |
There was a problem hiding this comment.
The starts/ends shortcut returns max(end)-min(start)+1 as soon as any phase has explicit weeks, which undercounts duration for mixed inputs where other phases omit start_week/end_week and are scheduled sequentially in the Resource Plan. In that case Summary can report a shorter calendar duration than the generated plan (e.g., one explicit phase W1–W2 plus one implicit 2-week phase still shows 2). This breaks the new calendar-duration metric and should fall back to rp_info["total_weeks"] unless every phase has explicit bounds.
Useful? React with 👍 / 👎.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Refactor of the
pmo-pert-estimateskill: the generated workbook is now a4-sheet PMI-compliant deliverable — WBS · Resource Plan
(Pianificazione Risorse) · Risks (Rischi) · Summary (Riepilogo) — with
PD as the single effort unit, correct Management Reserve base, native
PM/DevOps overhead, calendar-aware duration, and i18n-ready labels.
Why
Five issues observed in real-world usage drove the refactor:
meaningless Summary rollups (e.g. "Effort by Team = 2400" from
summing percentages).
Total PERT Durationsummed leaf durations sequentially →over-stated calendar duration by 2–3× when phases ran in parallel.
Contingency × pct, not(Tech + Overhead + Contingency) × pct→ violates PMI PMBOK §4.3.no real parallelism, misleading for clients.
them as fake WPs.
The 4-sheet layout, the new
Resource Plan(role × week PD matrix), andthe correct MR base resolve all five issues at the root.
Key changes
i18n.py(en/it labels),config_compat.py(legacyJSON detection + soft backfill with stderr warning),
wb_pianificazione_risorse.py(role × week PD matrix).wb_summary.py(Fascia BASSA/MEDIA/ALTA, calendarduration, effort by team via WBS primary role, sensitivity scenarios),
wb_risks.py(MR formula onWBS!H{total} × (1 + pm_pct + devops_pct) + Contingency).pm_overhead_pct,devops_overhead_pct,alta_uplift_pct,calendar_total_weeks,phase.start_week/end_week, top-levelscenarios[].wb_timeline.py,wb_resources.py, and their testmodules.
workbook with sensible defaults and a single stderr warning.
references/template-criteria.md rewritten for the 4-sheet layout.
anonymized
scenario-6-pa-portal-integration(Italian, 8 phases,parallel build/integrations, 20% MR) added.
assets/pert-template.xlsxregenerated from thev2
sample-input.json.Test plan
bash tests/ci/run-structural-tests.sh)python3 generate_excel.py --input examples/sample-input.json --output /tmp/x.xlsxproduces 4 sheets in canonical orderPianificazione Risorse,Rischi,Riepilogo)validate_template.pyaccepts the regenerated template (valid: true, no errors)#REF!errors.xlsxin Excel/LibreOffice and verify Resource Plan totals match WBS!H{total} (±1 PD)