Prepared 2 October 2026. This is the receiving record for ten repaired Project Apps examples and their five existing Library subjects. Receiving change: website PR175. Paired retirement: Project Apps PR149. Both are proposed changes; publication remains pending.

Source and identities

Original implementations were reviewed at fa5a4a4819c8c5d25b413c7d1f3219aac2e81fc2. The receiving _data/library_apps.json records each original catalogue ID, name, path, full source revision and SHA-256, alongside its maintained source directory. These identities and hashes were independently checked against that revision; Git retains the full original source.

All old files below are under Project-web-apps/web_apps/; all destinations are under the website’s /library/apps/.

Original ID and file Library destination Subject
25 · intro_risk_matrix.html risk-matrix/ Decisions & trade-offs
30 · weighted-decision-matrix.html weighted-decision-matrix/ Decisions & trade-offs
29 · eisenhower-matrix.html eisenhower-matrix/ Decisions & trade-offs
27 · decision-latency.html decision-latency/ Delivery dynamics & feedback
42 · critical-path-spotter.html critical-path-spotter/ Delivery dynamics & feedback
6 · hs2_WBS_to_PBS.html wbs-to-pbs/ Project states & relationships
34 · Project_Data_standard_example.html project-data-standard/ Data, evidence & assurance
38 · white_space_analysis.html white-space-analysis/ Capabilities & futures
32 · service_trident.html service-trident/ Capabilities & futures
28 · stakeholder-mapper.html stakeholder-mapper/ Project states & relationships

What is retained and corrected

  • Risk matrix: keeps named risks, 1–5 controls, score product, bands, grid and materials example. Multiple risks now survive in one cell; 3 × 5 is correctly Moderate. Adds editing, a register and explicit ordinal-score limits.
  • Weighted decisions: keeps all three vendor examples, criteria/options, editable weights/ratings, raw totals and removal confirmations. Missing or invalid ratings pause ranking, all ties are shown, and normalised scores remain stable when weights are rescaled.
  • Eisenhower matrix: keeps task ratings, all four action rules, scatter map, task list and optional support link. A 2.5 boundary agrees with the ≥3 high-rating rule; overlapping tasks remain inspectable. Editing no longer retains an earlier blank-name validation error.
  • Decision latency: keeps eight tasks, five decision durations, every association, both charts and cross-highlighting. Explanations now distinguish association from causal delay; the supplied data cannot identify a critical path. Adds keyboard selection and a full table.
  • Critical path: keeps the seven-task network, every add-one control, task bars and dependencies. Validated forward/backward passes calculate float and multiple critical routes; cycles and invalid inputs fail explicitly. Adds duration editing, reset and zero-duration milestone marks.
  • Work to products: keeps every original work/product/participant record, metrics, phase views, trees and relationships. Distinct product IDs resolve collisions; three previously unreachable outputs are exposed. Legacy HS2 figures and the process-oriented left tree are clearly qualified.
  • Project data: keeps editable JSON, example, codelists, SLA/tolerance, six readouts, three visuals and report export. Allows early completion and forecast savings; checks CAPEX currencies, zero tolerance and provenance. Invalid input clears stale reports; invalid money cannot create comparable chart bars.
  • Capability white space: keeps six dimensions, all five numerical score vectors, six plots, automatic selection, entrant sliders, rankings and CSV. Fictional labels replace unsupported company assessments; scores are now editable. Empty-region boundaries and zero-area cases are corrected. The speculative Controls + Twin + Contract offering remains as a labelled hypothesis.
  • Service Trident: keeps all three need-to-service mappings, multi-level selection, suggestions, clear and summary copy. Native controls preserve focus; failed or stalled clipboard access reveals manual copy. The app describes its mapping as an authored conversation heuristic.
  • Stakeholder mapper: keeps power/interest ratings, four contact categories, plot and list. Coincident entries remain distinct, thresholds and axes are explicit, and editing/removal preserve keyboard continuity. Judgements are not treated as measured support or a reason to exclude people.

The ten receiving apps retain their useful explanations and visualisations with local HTML/CSS/JavaScript, subject navigation, labelled controls and a cream/green Library frame. None of the ten originals used localStorage, sessionStorage or IndexedDB persistence; no stored browser data is cleared or migrated.

Catalogue and retirement scope

The old app-index.csv changes from 60 to 50 records, removing exactly these ten source identities and preserving unrelated rows. The receiving site adds ten static examples to its sixteen existing maintained apps: 26 Library apps in total. Both maintained manifests retain the old source names as lookup keys, point directly to the Library routes, and suppress duplicates when a stale pre-migration CSV is loaded. Independent execution of that stale-catalogue case produced 76 distinct identities: 50 general apps plus 26 maintained examples.

Only the ten old working implementations and their ten unused catalogue thumbnails are retired by this change. Each old HTML route becomes a tiny canonical bookmark forward that preserves query strings and fragments; it contains no duplicate app implementation. Prior move records remain unchanged, and source history remains available at the pinned revision. The static receiving apps are excluded from the earlier bundled-app rebuild loop so it cannot overwrite their maintained files.

Checks and evidence boundary

  • Decision regression group: 12 passed, covering products/bands, duplicate entries, weighted totals/ties/rescaling/invalid input and validation-to-edit recovery.
  • Schedule/breakdown regression group: 14 passed, covering CPM, supplied latency data, UTC dates, full tree reachability and cross-view identity.
  • Capability/data regression group: 12 passed, including a full-grid white-space oracle, funding-renderer invalid-currency cases and stalled-clipboard fallback. Stakeholder classification/range/orientation assertions also passed.
  • Independent integration review checked all ten source IDs/hashes, routes, subjects, manifests, stale-catalogue deduplication and deletion scope. Existing app-migration and earlier sixteen-app retirement checks passed.
  • Old repository’s complete node test.js passed, including all 23 canonical forwards with query/fragment preservation. Diff whitespace checks passed in both repositories.
  • Browser journeys passed for all ten apps, including shared risk cells and invalid-name-to-edit recovery, vendor ties and blank-rating pause, urgency editing, CPM float/finish changes, decision cross-highlighting, linked product inspection, early milestone validation, white-space boundary changes, service mapping/copy, and stakeholder overlap/edit/removal. All ten phone-width layouts fit the page at 390px; wider diagrams and tables scroll inside their panels. Ten diagram screenshots provide the pictured entrances. Exact Jekyll build and rendered-route checks run on PR175; its current check status is authoritative.
  • All 22 source Node checks passed; the prior sixteen apps passed 55 model tests and rebuilt to 49 unchanged files. Project-data report preview generation and download dispatch were observed. A completed browser download event and saved file have not yet been confirmed. No publication or human-use outcome is inferred from local tests.

Publication order

  1. Review and merge the receiving website change; build and publish its ten destinations.
  2. Verify the served destinations, subject entrances, dependencies and relevant interactions.
  3. Only then merge and publish the old-repository retirement; verify old links, query/fragment forwards and the 50-row catalogue.
  4. Record the actual review/deployment references and final browser/download outcomes here. At preparation time those publication references remain pending.