Labs

The app is designed as a local-first audit dashboard. It crawls public sitemap data, stores raw evidence locally, runs deterministic checks, and presents findings in a reviewable interface. AI-assisted observations can be included, but they are treated as candidates for human review rather than confirmed issues. Custom dictionary and false positive filters ensure a smooth, accurate run.

A typical workflow starts with entering a website, crawling its public pages, collecting page-level evidence, and surfacing issues such as missing metadata, content problems, accessibility concerns, or other repeatable checks. Findings can then be reviewed by the web team and exported to CSV or XLSX for sharing, prioritization, or follow-up automation.

The main value is operational clarity: Blackline Index turns website auditing into a repeatable internal process with evidence, exports, and a local dashboard built around the team's actual review workflow.

Upon initial run and false negative pass, over 1500 SEO, UX, and spelling issues were discovered across 31 sites in 2 hours. The team previously averaged 1 to 2 weeks per audit per site.

The system is built around a simple idea: AI is most useful when it has clear rules for what it should preserve, what it can summarize, and what it should leave untouched. Source material remains intact, synthesized notes are organized into durable knowledge, active projects stay separate from reference material, and private journal writing is protected from unwanted rewriting.

In practice, Command Vault turns scattered inputs into usable memory. Articles, notes, voice memos, project ideas, and reflections can be processed into concise summaries, linked to related concepts, and connected to active work without losing provenance. Each update leaves a maintenance trail, making the system inspectable rather than mysterious.

The result is a portfolio-scale example of human-AI knowledge management: a working environment where AI does not just answer questions, but helps maintain a living body of context with structure, privacy, and continuity.

The core design centered on a few durable rules. Tasks use four statuses: New, Active, Waiting, and Done, with completed work preserved in archive/search instead of cluttering the main dashboard. Attention is based on the last meaningful update, turning green, yellow, or red depending on freshness. Today pins are manual and persistent, giving the user a stable daily priority layer instead of an automatically shifting queue.

V1 also made room for the messy realities of the user's work. Email references are tracked manually by subject, sender, received date, and email status, avoiding premature Outlook integration. Store and brand scope is configurable, supporting individual stores, multiple stores, and brand-wide filters. The result is a dense, restrained app built around repeated daily use rather than a generic task system.

The value of Attention Ledger is its discipline: it keeps attention on what needs movement, preserves completed work, and avoids pulling the user into another overbuilt workspace. It succeeds by being narrow, local, and opinionated, turning a personalized Notion workflow into a faster dashboard that better matches how the user actually manages work.

The resulting Email Builder is a local campaign-production tool tailored to the client. It uses production-ready email templates, brand-specific guardrails, editable campaign fields, Hero Builder image generation, static QA checks, and browser rendering validation. For data-heavy campaigns, the tool can extract structured content from source URLs, including product pages, special offer pages, and new product launch pages for launch emails.

A key example is the Product Launch workflow. Instead of treating a URL only as a CTA destination, the builder can now read that URL, extract product variants, statistics, pricing, capability notes, and supporting launch copy, then feed those details into an approved email layout. The marketer still reviews and edits the content if needed, but the tool gives them a strong, factual first draft built from source material.

The result is a more reliable campaign workflow: faster production, fewer copy-and-paste errors, cleaner CTA strategy, stronger template consistency, and built-in checks for common email issues like missing alt text, weak CTA language, red styling, JavaScript, broken layout behavior, and Mailchimp merge-tag problems. The tool turns repeat email production from a manual design-and-QA exercise into a guided, brand-aware build process.

Early results showed a 40% lift in CTR and a reduction of production time from 4-10 days with a traditional agency to less than 45 minutes. A cost reduction of $200 in agency management fees per email was also observed.