Redesign the readiness report

[Gandy2025]

Summary

• redesign the Agent Team Readiness report around the approved Product Scan editorial direction
• keep the 0–100 score, six canonical dimensions, blocker ordering, evidence, minimum fix, and verification semantics source-bound
• make local and hosted machine-readable links honest and update the publishing recipe for the hosted contract
• add responsive, print, accessibility, escaping, withheld-score, and link-behavior coverage

Preview

Open the checked report example

Validation

• npm test — 52/52 passing
• npm run check
• git diff --check
• Playwright: 1440px desktop, 390px mobile, disclosure interaction, horizontal overflow, console, and full-page screenshots
• Letter PDF rendered and inspected page by page
• two independent final code reviews completed with no code blockers

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[Gandy2025]

@bestony The contributor reports completing the CLA as Gandy2025 from another device. I triggered the official CLA Assistant recheck endpoint afterward, but license/cla remains pending and the bot comment has not updated. The PR commit is authored and committed by the same linked GitHub account.

Could you please verify the CLA Assistant signature association/configuration or re-run the check from the maintainer side? This is now the only external status blocker; the repository test check is green.

[bestony]

Reviewed exact head d9002d136d6b211e0c1fca78565db632de20d585. The responsive presentation, escaping, print treatment, and local/hosted machine-link handling are sound, but the redesign drops two source-bound parts of the atr-1 contract from the human report.

Validation: npm test (52/52), npm run check, and git diff --check passed. I also rendered the checked example at 1440px and 390px without horizontal overflow.

license/cla remains pending; that is separate from this code verdict.

[bestony]

[P1] Preserve blockers outside the Top 3

This ledger is now the only per-dimension view, but it never renders dimension.must_fix_blocker. top_3_fixes is capped at three while the contract permits one blocker in each of six dimensions. Scanning fixtures/no-instructions produces five dimension blockers; the rendered HTML contains only the first three and silently drops the shared-context and handoff blockers. That changes report meaning rather than presentation. Please keep every dimension blocker and its evidence accessible, and add a regression using a report with more than three blockers.

[Gandy2025]

Fixed in ee73b21. Every dimension ledger detail now renders its must_fix_blocker with severity, impact, and complete blocker evidence, independently of the Top 3 chapters. The no-instructions regression asserts that all five dimension blockers render even though top_3_fixes contains only three. I also verified the two outside-Top-3 blockers interactively at 1440px and 390px and in the 8-page Letter print output.

[bestony]

[P2] Expose the weight used by the headline score

The redesign no longer renders dimension.weight anywhere. The contract uses unequal fixed weights and computes round(sum(score / 10 * weight)); readers cannot reconcile the displayed 0-100 headline with the six /10 values when those inputs are hidden. The previous renderer exposed each weight. Please restore the weights in the rail or ledger and assert all six in the renderer test.

[Gandy2025]

Fixed in ee73b21. The hero dimension rail now exposes each fixed weight beside its /10 score, and the renderer test asserts the exact name/weight pair for all six dimensions. The headline calculation and atr-1 data remain unchanged.

[bestony]

Maintainer-side CLA verification completed:

• I re-ran the official CLA Assistant recheck endpoint. license/cla is still pending and the bot comment is unchanged.
• GitHub resolves both the author and committer of d9002d136d6b211e0c1fca78565db632de20d585 to Gandy2025, so commit-to-account association is not the mismatch.
• The repository currently inherits an organization-level, shared CLA Assistant link to bestony/5a32ba8d5daa68b95adab6a9bba9c1cc. That Gist contains both ccla.md and icla.md, but CLA Assistant renders only the first non-metadata file (implementation). The current public signing page therefore serves the CCLA, which itself tells individual contributors to use the ICLA instead.

I cannot inspect CLA Assistant's authenticated signee table from this agent session, but the public configuration is not a correct individual-signing path. The maintainer fix is to link CLA Assistant to an ICLA-only Gist, keep CCLA authorization as a separate verified/manual path, and then have Gandy2025 personally sign the current ICLA version before rechecking. I have not manually overridden the legal status.

Separately, the exact-head code review found two report-contract regressions and has requested changes.

[bestony]

Reviewed exact head ee73b21ca21134a54534cd8a631145a56ea0993e.

Both prior findings are resolved: all six fixed weights are visible alongside their dimension scores, and every dimension blocker remains available with severity, impact, and complete evidence even when it falls outside the Top 3. The same blocker content is present in the print path.

Validation: npm test (52/52), npm run check, and git diff --check passed. A freshly generated five-blocker report retained all five blockers while preserving the deterministic Top 3, rendered without horizontal overflow or console errors at 1440px and 390px, and exposed all five blockers under print media.

Code verdict only: this exact head currently reports the test check as successful, but no license/cla check is present. This approval does not constitute CLA acceptance.