Website Review Report

Date: 2026-04-26
Scope: full content pass (index.html, blog/index.html, projects/index.html, projects/*/index.html, _posts/*.md)
Audience: executive recruiters and hiring managers
Agents run: 6 persona + 4 discipline (all fixed agents)

Executive Summary

• Full runbook execution is complete across all required persona and discipline reviewers.
• The strongest recurring risk is trust/credibility friction from inconsistent proof formatting and inconsistent writing metadata across surfaces.
• Structural SEO and accessibility gaps are concentrated in project detail delivery (JS-injected content, weak standalone /projects/ index, and semantic interaction patterns).
• Positioning is memorable, but role-fit is not explicit enough in first-pass recruiter/executive skim behavior.
• Contact and conversion paths are present, but not yet optimized for recruiter-specific next steps.

Categorized Findings

UX/Content

• Must: Writing metadata inconsistencies create avoidable trust damage.
  • Evidence: Homepage Writing entries do not fully align with post front matter (date/title mismatch for the AI post and title mismatch for the COS post).
  • Recommended fix: Source Writing card title/date from canonical post metadata (or manually enforce exact parity).
• Should: Role-fit clarity is delayed by personality-first framing above the fold.
  • Evidence: Hero/intro language is distinctive but not explicit about role scope in recruiter terms.
  • Recommended fix: Add one concise role/scope line near the hero and one compact “fit snapshot” block.
• Should: Contact copy is warm but not explicitly hiring-flow oriented.
  • Evidence: CTA is conversational and broad; no recruiter-targeted path framing.
  • Recommended fix: Add a hiring-specific CTA variant and clearly signal the preferred outreach path.
• Nice: Blog landing subtitle is implementation-centric instead of reader-value-centric.
  • Evidence: Subtitle describes stack/rendering rather than content value.
  • Recommended fix: Rewrite subtitle to express what operators/recruiters learn from the writing.

UI/Design

• Must: Keyboard focus visibility is inconsistent or absent on key controls.
  • Evidence: Hover states are present, but shared, explicit :focus-visible treatment is not consistently defined across nav, cards, links, and modal controls.
  • Recommended fix: Add one shared focus token system and apply it across all interactive components.
• Should: Mobile nav behavior appears better handled on homepage than shared templates.
  • Evidence: Homepage has tighter mobile nav handling while shared layout nav behavior is less robust on narrow widths.
  • Recommended fix: Centralize nav responsiveness in shared layout styles.
• Nice: Token discipline drifts in some template CSS via hardcoded colors.
  • Evidence: Literal color values are repeated in templates where semantic variables are preferred.
  • Recommended fix: Replace repeated literals with existing color variables for maintainability.

Code/Semantics

• Must: Homepage project cards use JS-upgraded list items instead of native link semantics.
  • Evidence: Cards are list items made clickable via JS with role="link" + tabIndex.
  • Recommended fix: Render cards as native <a> links (or links inside each card) and remove semantic polyfill behavior.
• Must: Project detail pages rely on runtime JS injection for primary content.
  • Evidence: Project files contain front matter only; critical case-study content is injected client-side.
  • Recommended fix: Server-render essential project content in static HTML, then progressively enhance with JS.
• Should: Landmark and heading semantics can be strengthened.
  • Evidence: Homepage lacks a clear <main> region; project detail heading hierarchy starts at visual styles not always semantic top-level headings.
  • Recommended fix: Add <main> around primary content and ensure each project detail has a true page-level heading structure.
• Should: Modal/lightbox focus flow is only partially handled.
  • Evidence: Open/close and escape handling exist, but robust focus trapping/return behavior is incomplete.
  • Recommended fix: Add focus trap and deterministic focus return to trigger element.

SEO/Accessibility

• Must: Crawl/discovery quality is constrained by thin JS-dependent project pages.
  • Evidence: Important project proof is not present as static, crawl-friendly HTML in project URLs.
  • Recommended fix: Publish static summaries and core proof on each project page; keep JS as enhancement.
• Should: /projects/ is redirect-only instead of a crawlable index hub.
  • Evidence: Immediate redirect shell with minimal fallback text.
  • Recommended fix: Replace with a lightweight projects index that links to every case study.
• Should: Metadata specificity should be expanded per project page.
  • Evidence: Project pages largely inherit generic site description.
  • Recommended fix: Add page-specific descriptions aligned to each project outcome domain.
• Nice: Homepage <title> can be more intent-aligned.
  • Evidence: Current title is minimal and omits role/value context.
  • Recommended fix: Extend title with role + operator context.

Persona/Resonance

• Must: Outcome proof consistency is uneven across projects.
  • Evidence: Some case studies are tightly quantified while others use broad qualitative impact labels in equivalent prominence.
  • Recommended fix: Normalize impact to bounded format: metric + timeframe + scope/ownership.
• Should: AI framing is strong but occasionally competes with operator identity.
  • Evidence: Repeated AI-forward cues can overshadow executive operator framing in quick scans.
  • Recommended fix: Keep AI as method-level proof, not the first identity cue.
• Should: A “coming soon” writing item in top slots can signal unfinished curation.
  • Evidence: Placeholder-style entry appears alongside published posts.
  • Recommended fix: Reorder published items first or relabel placeholder as draft-in-progress.

Prioritized Action Plan

Must (ordered)

1. Resolve writing metadata inconsistencies across homepage/blog/post surfaces.
   • Why now: This is the fastest trust recovery win and directly affects reviewer confidence.
   • Implementation notes:
   • Make writing card title/date derive from post front matter.
   • Align existing mismatched title/date strings immediately.
   • Add a quick metadata parity check before publish.
2. Convert homepage project cards to native link semantics.
   • Why now: This is a major usability/accessibility and semantic correctness issue in core navigation.
   • Implementation notes:
   • Replace JS role="link"/tabIndex approach with native anchors.
   • Preserve current visual layout and hover treatments.
   • Retain keyboard/assistive behavior through native HTML, not polyfills.
3. Make project detail pages static-first instead of JS-content-only.
   • Why now: Crawlability and direct-link credibility depend on server-rendered proof.
   • Implementation notes:
   • Render project summary, ownership, and impact in static HTML per page.
   • Keep enhanced modules (lightbox/interactions) as progressive enhancement.
   • Add page-level metadata and heading semantics while refactoring.
4. Add a shared focus-visible and modal focus-management accessibility baseline.
   • Why now: Keyboard navigation quality is a foundational accessibility blocker.
   • Implementation notes:
   • Define one reusable :focus-visible pattern for all interactive components.
   • Trap focus in modal/lightbox contexts and return focus on close.
   • Verify keyboard flow on homepage, blog, and project templates.
5. Standardize impact proof formatting across all featured projects.
   • Why now: Inconsistent proof rigor undermines overall credibility for executive/recruiter audiences.
   • Implementation notes:
   • Rewrite each project impact to include metric/timeframe/scope where possible.
   • Where exact numbers are unavailable, use bounded directional phrasing.
   • Apply the same proof template to card-level summaries and project headers.

Should (ordered)

1. Add explicit role/scope fit language above the fold and a concise recruiter skim block.
2. Add hiring-oriented CTA copy and a clearer primary outreach path.
3. Replace redirect-only /projects/ with a lightweight, crawlable project index page.
4. Unify shared mobile nav behavior across homepage, blog, post, and project layouts.
5. Soften abstract or inflated phrases in high-visibility sections to match tone guide precision.

Nice (ordered)

1. Rewrite blog subtitle to audience value, not implementation details.
2. Reorder or relabel placeholder writing entries to avoid unfinished signaling.
3. Clean minor copy polish issues (spacing/punctuation and small text consistency).

Notes on Deduping

• Exact duplicate findings and same-root-cause repeats were merged.
• Distinct root causes were kept separate even when related (for example, role-fit clarity vs proof standardization).
• This report intentionally preserves thoroughness while removing redundancy across all 10 reviewers.