Logo
Practices
Offerings
AI Bootcamp
Industries
Work
About
Insights
The Agentic Enterprise
Contact
Logo

Spearhead V6.3 Detailed Red Team Report

October 11, 2026 · Website and Standards · Complete local colleague review.

V6.3 is built as a 55-page website review with four practice pages, fourteen offering review pages, four implementation directions, four preserved capability entry points, five selected Bootcamp variants, substantive industry/company/delivery pages, three source-qualified cases, editorial examples and complete local utility journeys. The current hero and visual identity are retained. Newsletter-derived dividers, contrast and reading rhythm are applied to the homepage and detailed layouts.

The completed candidate is suitable for colleague review. No unresolved critical, high or medium build defect remains within the tested scope. Public release remains subject to the commercial definitions, evidence permissions and launch dependencies listed below. This report does not certify whole-site accessibility, regulatory compliance, commercial readiness or performance in every browser.

Plan confirmation and completion

The detailed October 11 plan was the build scope. Four practice pages and all fourteen offering items have built review destinations. Sauce, AI Forward and Governance as a Service are in a separate concept/service-design review directory, outside public offer recommendations. Legal, HR, Supply Chain and Tech Support are bounded implementation directions, not claims of four delivered platforms. Five selected ABC variants are grounded in the available catalog and collateral inventory; the sketch does not establish a compulsory 22-page website.

Step Completion Evidence
Portfolio and page inventory Complete source/portfolio-v6.3.json; source/page-register.json; all fourteen items accounted for.
Isolated V6.3 build and structure Complete V6.2 remains separate; four practices use updated names; FinOps belongs to Strategy and Transformation.
Detailed practice and offering pages Complete for review Four practice pages; fourteen offering pages; original copy, finished visuals, scope, outputs, examples, FAQs and next steps.
Implementation, capability and supporting pages Complete for review Four workflow directions, four capabilities, five ABC variants; full case, editorial, company, delivery and utility paths.
Homepage and customer journeys Complete Homepage readability pass; corrected Studio, Bootcamp, downloads and inquiry carryover.
Red team and corrective loop Complete Three independent reviewer perspectives plus parent source, browser, contrast, layout, interaction and package checks.
Standards and portable review package Complete Reconciled standards, page register, this report, reviewer evidence and extracted-package verification.

Review dimensions and method

Three separate reviewers examined visuals/readability, commercial evidence/taxonomy, and copy/buyer journeys. They reviewed work authored outside their own copy task and reported findings while the build was changing. Their detailed reports retain initial observations and final dispositions. The parent consolidated duplicate findings, fixed the build, ran the full browser matrix and rechecked affected behavior. This is an adversarial multi-perspective review, not a formal external audit.

Dimension Review and evidence Result and limits
Operating structure Current October 11 inventory against practice/offer grouping and page coverage. Four correct practices, fourteen accounted items, no invented practice heads or extra register items.
Offering substance Buyer trigger, outputs, scope, responsibility, acceptance and boundary across all detailed copy. Discovery/Blueprinting overlap explicit; factory design versus implementation separate; existing-pilot and architecture paths preserved.
Evidence and claims Catalog and working-draft maturity against page promises; existing case claims compared with V6.2. No invented new client results, fixed prices/durations or delivered Software Factory case. Case evidence states retained.
Copy and cognition Natural prose, specific examples, headings, repeated scope language and action labels. Short action labels and homepage summaries; distinct detailed copy. A low editorial watch item remains for shared structural rhythm and repeated scope cautions.
Visual hierarchy and brand Four generated editorial practice maps; native diagrams; current desktop/mobile panels. Existing palette/type/hero retained; diagrams now express relevant relationships; no production-interface or real-workshop-photo impersonation.
Readability and mobile 55 pages at 1440, 1024, 768, 390 and 360px, with screenshot evidence. 275 checks: no viewport overflow, image failure, duplicate ID, page script error, unlabeled form field or visible native SVG text clipping.
Accessibility Semantic headings/labels/skip links, selected keyboard flows, reduced motion, equivalent-width reflow and computed text contrast samples. Recorded tests pass after fixing a white-on-white Bootcamp action. This is not a screen-reader, manual WCAG audit or complete accessibility certification.
Journeys and function Every Studio situation × every support choice; format/query/change/pageshow/Enter, dialog and draft download tests. 112 Studio combinations and 16 main journey checks pass; additional targeted accessibility interactions pass.
Privacy and security Plain-text query prefills, known route/topic validation, escaped constant labels, no outbound form submission/storage/analytics. Local drafts remain unsent; conceptual demo executes no AI action. No backend, authentication or hosted-service security test is claimed.
Performance and assets Locally bundled fonts/assets and WebP practice illustration sizes; layout/image loads. Four new practice raster assets total about 532 KB. No external requests in page matrix. No network-throttled Core Web Vitals score or production SLA is claimed.
Delivery and packaging Current version records, working/review consistency, local file paths and extracted offline workflows. Portable review candidate includes the same final website. Publication and the older PowerPoint library are separate.

Findings and corrective actions

The initial defects below were observed during source and browser reviews. Severity describes their initial consequence. Each is corrected in the final candidate; the evidence states how it was rechecked. Reviewer reports sometimes retain “pending” notes from earlier snapshots; these final consolidated dispositions govern the delivered candidate.

F01 High FAR falsely appeared as a sequence

Observed: The initial corridor and numbered phone view implied Current work → Fix → Accelerate → Reimagine. Copy and source describe options.

Corrected: Replaced it with three branches, option labels and an explicit alternatives description.

Final disposition: Closed for this local build. work-redesign.html; diagrams.py; independent visual/commercial recheck.

F02 High Some requested pages could not build

Observed: Null practice fields reached the shared template and caused a rendering exception.

Corrected: Normalized page ownership and gave company/industry pages purpose-specific breadcrumbs and labels.

Final disposition: Closed for this local build. 55 emitted pages; 275 browser loads.

F03 High Inquiry script failed to parse after repeated repair

Observed: Repeated insertion duplicated a lexical declaration and disabled prefills and draft generation.

Corrected: Regenerate from the pristine baseline; syntax-check scripts and exercise actual draft creation/download.

Final disposition: Closed for this local build. inquiry.js syntax pass; downloaded inquiry-draft-example.txt.

F04 High Leaders could revert to a one-day outline

Observed: A separate format patch missed Enter/pageshow and lost its mutable label matches.

Corrected: Integrated deterministic cohort behavior in every shared render, with a separate leadership action/prefill.

Final disposition: Closed for this local build. All five formats through query/change/pageshow; Enter and return-to-Executive tests pass.

F05 High Bootcamp inquiry action became white on white

Observed: The legacy component referenced --blue while the new detailed-page stylesheet defined --d-blue.

Corrected: Provide explicit compatible tokens for the shared component; rerun contrast checks.

Final disposition: Closed for this local build. 55-page computed-color inspection: no failing sampled pairs after correction.

F06 Medium Value questions reached the wrong offer

Observed: Strategy/value fell through to Discovery; original FinOps ownership was Governance.

Corrected: Route value and cost choices to Value and AI FinOps under Strategy and Transformation.

Final disposition: Closed for this local build. 112 route/support combinations check destination, practice and brief.

F07 Medium Unselected opportunities reached Blueprinting

Observed: Unsure/opportunity and readiness recommended work before the problem was selected.

Corrected: Both now reach Discovery; a selected engineering idea can still reach Blueprinting.

Final disposition: Closed for this local build. All eight affected support combinations rechecked successfully.

F08 Medium The contact handoff omitted the plan

Observed: The visible promise exceeded a URL containing only title/practice.

Corrected: Carry the full deterministic brief as plain text into the editable inquiry.

Final disposition: Closed for this local build. Download and brief parameter equality on all 112 routes; actual inquiry/download test passes.

F09 Medium Repeated builds duplicated Studio choices

Observed: Repair insertions accumulated repeated engineering situations and mapping branches.

Corrected: Generate homepage, Studio and inquiry from pristine V6.2 input on each build.

Final disposition: Closed for this local build. Situation IDs are unique; one Engineering Operating Model option.

F10 Medium General inquiries inherited Engineering intent

Observed: About and industry links prefilled a production-system need the reader had not selected.

Corrected: Use neutral general inquiry context on company/industry pages.

Final disposition: Closed for this local build. About prefill test confirms no production-engineering assertion.

F11 Medium The configurator did not identify function

Observed: Audience/workflow choices did not capture which function the program served.

Corrected: Add a distinct function choice and carry it into the outline and inquiry.

Final disposition: Closed for this local build. Leaders handoff retains Function: Customer support; six detailed choices, five condensed homepage choices.

F12 Medium Nonlab programs promised agent building

Observed: The shared preview and condensed homepage assumed all formats built agents.

Corrected: Use agreed practice/output language for nonlab formats and the homepage; retain lab wording for Agent lab/Hackathon.

Final disposition: Closed for this local build. Format tests pass; final homepage wording and dynamic template verified by independent reviewer.

F13 Medium Paragraphs were rendered as action labels

Observed: Long invitations created large multilined buttons on phones.

Corrected: Use concise action labels and preserve the invitation as adjacent closing prose.

Final disposition: Closed for this local build. Independent targeted desktop/390/360 inspection and parent gallery review.

F14 Medium Homepage row descriptions became too dense

Observed: Using full decks increased the practice section and reading burden.

Corrected: Write separate concise summaries for homepage rows; keep detail-page decks substantive.

Final disposition: Closed for this local build. Independent before/after geometry and mobile panels.

F15 Medium Every diagram used the same relationship

Observed: Initial four-pad graphics did not distinguish choices, layers, controls, feedback or human review.

Corrected: Built branching, layer, cost/value convergence, decision, assistance/review and feedback patterns with page-specific labels.

Final disposition: Closed for this local build. Native diagram inspection and 275 visible-label/geometry checks.

F16 Medium Native explanations dropped control/feedback

Observed: Practice captions promised equivalent meaning while human control, client authority or feedback was missing.

Corrected: Add the governing relationships to live descriptions, captions and phone equivalents; improve image alt text.

Final disposition: Closed for this local build. Four practice illustrations paired with accessible native descriptions.

F17 Medium Tablet diagram labels became too small

Observed: 16px source labels shrank in a 960px SVG at tablet width.

Corrected: Use a readable text equivalent through 1000px and larger desktop labels.

Final disposition: Closed for this local build. 1024/768/390/360 layouts, SVG bounds and native text inspected.

F18 Medium Bootcamp variants repeated too much copy

Observed: Several variants shared all preparation/output/FAQ substance.

Corrected: Differentiate openings, source preparation, tangible outputs and setting-specific questions; retain common scope limits.

Final disposition: Closed for this local build. Commercial reviewer confirms distinct variant content and bounded asset audit.

F19 Medium Case presentation contradicted evidence

Observed: A shared example disclaimer treated reported case sequences as invented; one label implied ongoing portfolio operation.

Corrected: Use bespoke reported-case layouts, source qualifications and Reported production use wording.

Final disposition: Closed for this local build. Three case pages retain exact evidence states and anonymous attribution.

F20 Medium Internal scaffolding leaked into buyer copy

Observed: About and program boundaries referred to approved profile selections and the operating sketch’s page counts.

Corrected: Replace it with direct biographies and public scope language; keep provenance and collateral counts in review records.

Final disposition: Closed for this local build. Independent emitted-page copy review.

F21 Medium Validation and version records were stale

Observed: Copied V6.2 records and V8/V9 labels could be mistaken for current QA.

Corrected: Remove inherited QA, regenerate evidence and align active visible labels to V6.3/October 11.

Final disposition: Closed for this local build. Fresh geometry, Studio, journey and accessibility JSON; package metadata sweep.

F22 Medium Register provenance was too generic

Observed: One source description could not distinguish concepts, cases, biographies and variants.

Corrected: Record specific source files/types and the bounded Bootcamp audit; keep publication status separate from built status.

Final disposition: Closed for this local build. 55 page register entries with source_files and source_type.

Verification records

• 275 page and width checks: 55 website pages × five widths. Fresh V6.3 evidence; inherited V6.2 results were removed.

• 112 Studio checks: seven priorities, 28 situations and four support choices. Destinations, practice ownership, download substance and full-brief inquiry handoff checked.

• 16 main journey checks: actual draft/download, all five formats, selected query/change/pageshow/Enter cases, neutral inquiry intent, illustration dialog and all engineering scenarios.

• Accessibility checks: 55-page computed-color samples plus keyboard, reduced motion, reflow and targeted final wording checks. Color checks skip image/gradient backgrounds and do not exhaust hover/focus states.

• Page register: all 55 destinations with build/readiness/source status.

• Bootcamp asset audit: selected evidence and inspection limits.

• Independent visual review, commercial review, and copy/journey review: full findings and final dispositions.

• Current full-page desktop/phone captures are retained in the working build’s review/screens directory. The portable package omits this large screenshot set; review conclusions and machine-readable evidence remain included.

The automated image check loads lazy assets explicitly. Hidden/offscreen lazy images flagged in the first probe were a measurement limitation, not missing assets. The revised probe decoded the referenced local images and passed. A first route run exposed eight unsure-priority mismatches; the corrected run passes all 112. Initial raw “undefined” text flags matched ordinary prose, which was simplified; no undefined runtime values were displayed.

Publication dependencies and low editorial watch item

These conditions do not prevent a complete review build. They do prevent treating every page as a settled commercial promise or the package as a launched website.

Dependency Required decision or evidence Current treatment
Sauce First user/workflow, product boundary, tenancy/access, evaluation, support and funded ownership. Detailed concept review only; no price, subscription or ready-to-buy recommendation.
AI Forward Audience, distinct outcome, reinforcement, scope and difference from AI for Leaders. Detailed program-design review only.
Governance as a Service Covered work/volume, client authority, cadence, service expectations, exclusions and staffed capacity. Separate service-design review; no public availability promise.
Implementation directions Defined workflow, data/integration perimeter, evaluations, operation and delivery readiness. Four substantive illustrative directions; no invented delivered platform claim.
FAR and Discovery boundaries Detailed FAR definitions/examples and independent Discovery versus included blueprint discovery. Proposed FAR explanation and explicit scope boundaries.
Case/logo permissions and proof Confirm public usage/labels, customer/partner/prior-employer distinctions, claim periods and approvals. Existing logos retained for local review; anonymous cases have evidence qualifications. No endorsement approval is inferred.
Contact/newsletter and hosted launch Production destinations, privacy terms, subscription/contact integration and publication decision. Local plans/inquiries are explicitly unsent; signup is not connected.
Further ABC variants Distinct audience/question and source/asset coverage. Five useful selected variants; identified PPT/PDF assets were not fully re-audited slide by slide.

The copy reviewer’s low watch item concerns repeated section rhythm and necessary scope cautions across related offerings. The shared layout is intentional for navigation consistency. Variant-specific openings, examples, outputs and questions were strengthened; material scope/maturity disclosures were retained. Colleague feedback can guide further editorial compression without removing those distinctions.

Practical limits

Browser testing used Chrome on macOS. Safari, Firefox, Windows, assistive screen readers, delivered email clients, production network conditions and hosted backends were not separately tested. Reflow checks included an equivalent 720px content width for a 1440px display at 200%; this is not a complete browser-zoom/assistive-technology audit. No claim is made that each qualitative category achieved an objectively measured 9/10. The original plan’s 9/10 goals guided critique; concrete checks and corrected findings support readiness for review.

V6.3 is the latest local website-and-standards review. The September V8 web-and-PowerPoint library is older and separate. The last explicitly approved September homepage snapshot remains preserved. This build does not overwrite an approval or claim publication.

Final extracted package verification

The extracted archive passes all 67 standalone HTML-document checks at phone width, with no overflow, broken images, script errors or external network requests. The email fragment is excluded from standalone-page heading checks; the superseded authoring demo remains in V6.2.

Offline plan carryover, feedback/download, inert query input, invalid-route handling, enlarged reading and the standard-font preview pass. Read the final package evidence →

Copy review follow-up

October 11, 2026. Reviewed all 55 website pages: headings, body copy, outputs, diagram labels and captions, FAQs, buttons, form labels and interactive plan text. Corrected 48 website HTML files and one JavaScript file, with matching authoring copy and shared-template updates.

  • Replaced “Agree the boundaries” with “Agree on the scope” on all 35 affected pages.
  • Corrected similar “agree” constructions and used clearer diagram labels: “Decide next steps”, “Choose next action” and “Confirm the scope”.
  • Corrected missing terms and awkward sentences, including acceptance criteria, an unspecified dependency, and the Blueprinting implementation decision.
  • Clarified dense wording while preserving offering definitions, concept statuses and qualified case figures.

Review without an app server

Extract the complete ZIP and open Start here.html in a desktop browser. All 55 website pages, standards, plans, inquiry drafts and feedback work locally. No app server, installation, Codex or account is needed. Enable JavaScript for the interactive guides. Inquiry and feedback forms prepare drafts; they do not send them.

Final extracted-ZIP offline checks →
Specific corrections and changed files →

The extracted ZIP passed 335 page-and-width checks across all 67 standalone HTML documents, including all 55 website pages. Six local interaction checks passed. There were no attempted external requests, failed local assets or JavaScript errors.

Earlier red-team observations remain as historical evidence. The revised website copy is the current candidate.

© 2026 Spearhead. All rights reserved.

Linkedin