🧰 Engineering practice

Checks and fixes how AI agents and answer engines reach a site, from robots.txt groups and Content-Signal per AI crawler to llms.txt, Markdown versions of pages, /.well-known discovery files, real 404s, accessible names and Lighthouse's Agentic Browsing score. Use when asked whether AI agents can read or use a site, about llms.txt, Content-Signal, ai-catalog.json, WebMCP or agent-friendly pages, or when seo-aeo-audit reports agent findings.

Diagnoses bugs by reproducing first and naming the root cause before any code changes — reproduce, instrument, bisect, prove the fix against the reproduction, leave a regression test. Use whenever something is reported broken, failing, or behaving wrong, before attempting a fix.

Interrogates a finished diff adversarially before it becomes a commit or PR — correctness, edge cases, security, and performance challenges the author must answer or fix. Use when a feature or fix is declared done, before committing, opening a PR, or handing work to review.

Enforces Uplift Duo house code style — naming, comments, error handling, and structure. Use when writing, editing, or reviewing code in any language.

Produces dual effort estimates — traditional human-developer weeks alongside an agent-driven build-days + tokens + cost figure assuming a tight, increasingly-unsupervised feedback loop. Use when asked to estimate effort, time, cost, timeline, "how long", "how much", or scope for a build, feature, migration, or refactor.

Applies the Uplift Duo git workflow — branch naming, commit triggers, the emoji commit message format, GitHub defaults, and merge rules. Use when branching, committing, merging, or creating repos.

Checks and fixes how a local business shows up in local and map search, from its Google Business Profile and name, address and phone consistency to LocalBusiness structured data, location and service pages and reviews. Use only when the site belongs to a business with a physical location customers visit or a service area it travels to (clinics, shops, restaurants, trades, law and dental offices); skip it for online-only businesses, SaaS and publishers.

Write and verify an AGENTS.md so any coding agent knows how to build, test, and stay inside your rules.

Review a diff for bugs that pass CI but break production, ranked with file/line evidence and missing tests.

Challenge a feature request, find the real job behind it, and cut it to a one-week version with metrics.

Audits a directory of git repositories and produces an interactive Project Audit — per-repo descriptions, % complete, ratings (market size, value, code quality, security, momentum), bugs by severity, feature suggestions and improvements with type tags, major-change timelines, tech-stack analysis with recommended alternatives, AI model attribution by percent, token usage (commit-block estimates vs measured transcripts with a completeness comparison), total AI thinking time, and a copy-paste improvement prompt per repo. Use when the user asks to audit, review, rate, or take stock of their repos, projects, or portfolio, asks "what's the state of my projects", or wants a project audit for any time window.

Reads how a site really performs in search from first-party sources (Google Search Console, Bing Webmaster Tools, Chrome UX Report field data and PageSpeed Insights) and turns it into a short list of fixes. Use when asked about rankings, clicks, impressions, indexing, Core Web Vitals, or whether an SEO change worked, or when seo-aeo-audit passes but pages still aren't found.

Creates and verifies the search, answer-engine and agent setup for a site's public pages, from titles, canonicals and structured data to sitemaps, robots rules, llms.txt and the files AI agents look for, then gates it on checks that catch real failures. Use when adding or auditing SEO, AEO or GEO, when a site publishes pages from a database or library, or when new pages should appear in search automatically.

Produce contracts, failure modes, a test matrix and 2-hour tasks for a scoped feature before coding starts.