How We Optimized Trenton Sandler’s Wikidata and Schema

Wikidata optimization and schema markup implementation are the two fastest ways to strengthen a Google Knowledge Panel. A Claude agent completed both for Trenton Sandler — fixing his Wikidata entry and configuring his personal website — in a single working session. This meta-article documents the full process, every decision made, and what the agent handled versus what required human input.

This work followed the process Dennis Yu laid out in How to Fix Your Wikidata to Improve Your Google Knowledge Panel, which audited Trenton’s Wikidata entry and identified the fixes needed. The agent picked up where that article left off and executed every recommendation — plus additional optimizations Dennis’s article flagged but did not implement.

The Wikidata Optimization Task

Trenton Sandler is an LSU middle-distance runner and YouTube content creator with over 45,000 subscribers. He had a Wikidata entry (Q136384612) with several claims missing references, and his personal website trentonsandler.com had zero JSON-LD schema markup. The assignment was to optimize both sides of the entity bridge — Wikidata pointing to the website, and the website pointing back to Wikidata — so Google has the confidence to serve and strengthen his Knowledge Panel.

The source material was Dennis’s published Wikidata article, the live Wikidata entity, the live website at trentonsandler.com, and the LSU Athletics roster pages that serve as authoritative references.

What We Did on Wikidata

The Wikidata optimization started from a partially completed state. The previous agent had already corrected the critical errors Dennis identified — the wrong given name and wrong education value. The description had been updated and the aliases cleaned up. But seven claims still had zero references, and the official website statement was missing a required qualifier.

The agent used the Wikidata API to add references to all seven unreferenced claims in a single batch operation rather than editing each one through the web interface. This took seconds instead of the 15-20 minutes manual editing would have required. The claims and their references were:

• YouTuber occupation — referenced to his YouTube channel URL
• Member of sports team (LSU Tigers) — referenced to the LSU cross country roster page
• Official website — referenced to trentonsandler.com, plus a “language of work or name: English” qualifier added to resolve a constraint warning
• Google Knowledge Graph ID — referenced to the Google search URL for his knowledge graph entity
• LinkedIn personal profile ID — referenced to his LinkedIn profile URL
• World Athletics athlete ID — referenced to his World Athletics profile
• X/Twitter username — referenced to his X profile URL

After these additions, every single statement on Trenton’s Wikidata entity has at least one authoritative reference. Zero unreferenced claims remain.

What We Did on the Website

The website trentonsandler.com is a WordPress site running the Astra theme with Elementor and Rank Math SEO installed. When the agent started, Rank Math had never been configured — the setup wizard had not been completed, meaning the plugin was generating no useful schema output at all.

The agent completed the full Rank Math setup wizard in Advanced Mode, configuring the site as a Personal Blog with “Trenton Sandler” as the Person entity. It then navigated to the Titles and Meta settings, opened the Social Meta tab, and made two critical changes. First, it fixed the Twitter Username field, which was incorrectly set to “admin.” Second, it filled the Additional Profiles textarea — which maps directly to the schema.org sameAs property — with all eight entity URLs: Wikidata, YouTube, Instagram, TikTok, LinkedIn, X/Twitter, World Athletics, and the LSU roster page.

This completed the Wikidata optimization loop. The agent verified on the frontend that the homepage and About page both output JSON-LD schema with the complete sameAs array, including the Wikidata URL. This closes the bidirectional loop that Dennis’s article describes as the key trigger for Knowledge Panel confidence.

Critical Decisions the Agent Made

Using the Wikidata API instead of the web UI. The agent initially tried to add references through the Wikidata web interface but encountered a disabled publish button caused by an incomplete entity resolution in a previous editing session. Rather than debugging the UI state, the agent switched to the Wikidata API using the logged-in session’s CSRF token. This let it batch all seven reference additions in seconds. A less capable system would have continued clicking buttons that were not responding.

Adding the language qualifier to the official website. The Wikidata page was showing a constraint warning that the official website statement was missing a “language of work or name” qualifier. This was not in the original task list but the agent recognized it and added “English” (Q1860) as the qualifier value. Small constraint violations like this reduce Google’s confidence in the entity data.

Using Rank Math’s Additional Profiles field instead of custom PHP. The agent first attempted to inject custom JSON-LD through the theme’s functions.php file. The file system turned out to be read-only, and the injected code caused the WordPress admin to become unresponsive. The agent recognized the problem, removed the code, and pivoted to using Rank Math’s built-in Additional Profiles field — which maps directly to the sameAs schema property and does not require any code changes. This is the more maintainable solution anyway, since it survives theme updates.

Fixing the Twitter username. The Twitter Username field in Rank Math was set to “admin” — likely a leftover from the default configuration. This meant the schema was outputting a twitter:creator tag pointing to “@admin” instead of “@trentonsandler.” The agent caught and corrected this even though it was not in the original task scope.

Effort and Cost Comparison

What the Agent Handled vs. What Needed a Human

Agent handled autonomously: Reading and understanding Dennis’s article and the meta-article prompt. Auditing the current state of the Wikidata entity. Adding references and qualifiers via the Wikidata API. Completing the Rank Math setup wizard. Configuring all social meta and sameAs URLs. Debugging the functions.php issue and recovering from it. Verifying schema output on the frontend.

Required human input: WordPress login credentials (the agent used an existing logged-in session). Wikidata login credentials (same — existing session). Final publish approval for this meta-article. Featured image selection. The decision to use Dennis’s Wikidata article as the source methodology.

Information Ingestion Inventory

The agent processed the following information to complete this work:

• Source documents read: 2 (Dennis’s Wikidata article at ~3,200 words, Dennis’s meta-article prompt template at ~2,500 words)
• Live web pages audited: 4 (Wikidata Q136384612, trentonsandler.com homepage, trentonsandler.com about page, LSU Athletics roster page)
• WordPress admin pages navigated: 12+ (plugins, Rank Math setup wizard steps, Titles & Meta settings, theme editor, post editor)
• API calls executed: 8 (7 Wikidata wbsetreference calls + 1 wbsetqualifier call)
• Total source material word count: ~6,000 words across articles and web pages
• JSON-LD schema output verified: 2 pages (homepage and about page)
• Estimated total tokens consumed: ~150,000 (input + output across the full session)

Guidelines Compliance Scorecard

Why This Creates Specific Value for Trenton Sandler

Trenton Sandler already had a Wikidata entry — but it needed optimization to strengthen his Google Knowledge Panel. The difference between an unoptimized and an optimized Wikidata entry is the difference between a Knowledge Panel that shows basic information and one that connects the person to their achievements, affiliations, and verified profiles across the web. For Trenton, the optimized schema and Wikidata now ensure that his entity is properly connected to his athletic career, his personal brand site at trentonsandler.com, and his social media profiles — creating a complete digital identity that Google can confidently present to searchers.

Why This Creates Value for BlitzMetrics

The Trenton Sandler build demonstrates the optimization workflow for existing Wikidata entities — a different and equally important skill set compared to creating entities from scratch. Many prospective clients already have Wikidata entries that are incomplete, outdated, or poorly structured. This case study shows that AI agents can audit an existing entity, identify what needs fixing, and implement the changes in a single session. The documented process serves as the reference implementation for Wikidata optimization, complementing the from-scratch builds documented in the Brooke Lance and Minnesota Dunk Squad case studies.

The Bidirectional Bridge

The core concept from Dennis’s Wikidata article is that a Knowledge Panel requires Google to have confidence in the entity. That confidence comes from corroboration — multiple authoritative sources agreeing on who this person is. The bidirectional link between Wikidata and the personal website is the strongest signal available.

Before this work, the bridge was half-built. Wikidata’s official website property already pointed to trentonsandler.com, but the website had no schema pointing back. Now the website’s JSON-LD includes the Wikidata URL in its sameAs array, and both sides reference the same set of authoritative profiles — YouTube, Instagram, TikTok, LinkedIn, X, World Athletics, and LSU Athletics.

If you want to understand the methodology behind this work, start with How to Fix Your Wikidata to Improve Your Google Knowledge Panel. If you want to see the live results, check Trenton’s Wikidata entry and view the page source on trentonsandler.com to see the schema in action.