/entity-audit

name: entity-audit
description: "Audit brand entity consistency across the knowledge sources AI engines trust — Wikidata properties, Google Knowledge Panel, Wikipedia presence and notability, and industry directories — producing a consistency scorecard, per-property discrepancy list, and a fix plan ranked by AI-visibility impact. Audits and recommends; it does not edit those platforms for you. Triggers on \"/digital-marketing-pro:entity-audit\", \"is our Knowledge Panel accurate\", \"our Wikidata entry shows the wrong founding date\", \"audit our entity data\", \"why do AI engines get our company facts wrong\". Reads the brand profile as the source of truth and logs each finding via geo-tracker."

/digital-marketing-pro:entity-audit

Purpose

Audit brand entity data consistency across the platforms that AI engines use as knowledge sources. Check Wikidata entries, Google Knowledge Panel accuracy, Wikipedia presence and notability, and industry directory listings for consistency. Inconsistent entity data degrades AI engine trust and visibility — when knowledge sources disagree about basic facts like the official website, founding date, headquarters location, or industry classification, AI engines either omit the brand entirely or present conflicting information. This command provides a systematic, platform-by-platform audit with specific discrepancies flagged and a prioritized fix plan ordered by impact on AI visibility.

Input Required

The user must provide (or will be prompted for):

• **Brand/entity name**: The exact name of the brand, organization, person, or product to audit — must match the entity as it should appear in knowledge sources. If the brand has known aliases or former names, include those for cross-referencing
• **Entity type**: `Organization`, `Person`, `Product`, or `Brand` — determines which properties are checked and which directory types are relevant. Organizations check founding date, headquarters, industry; Products check manufacturer, launch date, category; Persons check role, affiliation, notable works
• **Key properties to verify**: Official website URL, founding date, headquarters location, social media profiles (LinkedIn, Twitter/X, Facebook, Instagram), industry classification, key people (CEO, founders), parent organization, number of employees, and any entity-specific properties the user considers critical. Properties from the brand profile are used as the source of truth
• **Directories to check (optional)**: Industry-specific directories (e.g., G2, Capterra, Clutch for SaaS; Yelp, TripAdvisor for hospitality), professional associations, and business registries relevant to the brand's industry. If not provided, the command will suggest directories based on the brand's industry classification from the profile

Process

1. **Load brand context**: Read `~/.claude-marketing/brands/_active-brand.json` for the active slug, then load `~/.claude-marketing/brands/{slug}/profile.json`. Extract the authoritative values for all entity properties — official name, website, founding date, headquarters, social profiles, industry, key people, and description. These become the source of truth against which all platforms are compared. Also check for guidelines at `~/.claude-marketing/brands/{slug}/guidelines/_manifest.json`. If no brand exists, ask: "Set up a brand first (/digital-marketing-pro:brand-setup)?" — or proceed with user-provided values. 2. **Check Wikidata**: Search for the entity on Wikidata by name and aliases. If found, verify each property — official website (P856), social media profiles (P2002, P2003, P2013, P4264), founding date (P571), headquarters (P159), industry (P452), key people (P169, P112), instance of (P31), and description. Record each property as matching, mismatched (with both values), outdated, or missing. If no Wikidata entry exists, record as absent and assess whether the entity meets notability criteria for creation. 3. **Check Google Knowledge Panel**: Verify Knowledge Panel existence for the brand name query. If present, check whether the panel is claimed or unclaimed, whether displayed information (website, address, social links, description, category) matches the brand profile, and whether images and logos are current. Record each element as accurate, inaccurate (with discrepancy details), outdated, or missing. Note the panel source attribution. 4. **Assess Wikipedia presence**: Search for the entity on Wikipedia. If an article exists, verify accuracy of key facts — founding date, headquarters, description, key people, products/services, and any claims that could be outdated or incorrect. Check for citation quality and recency. If no article exists, assess notability criteria — significant coverage in reliable independent sources, demonstrated importance in the field, and verifiable claims. Record as present-and-accurate, present-with-issues (list issues), or absent with notability assessment (likely notable, borderline, or unlikely notable). 5. **Check industry directories**: For each relevant directory, verify the listing exists and check data consistency — business name spelling, address, phone number, website URL, business description, category classification, and any directory-specific fields. Record each listing as consistent, inconsistent (with specific discrepancies), incomplete (missing fields), or absent. Flag NAP (Name, Address, Phone) inconsistencies specifically, as these have outsized impact on entity resolution by AI engines. 6. **Record findings**: Store each entity finding via geo-tracker's `entity-check` action (`--platform` takes an entity platform: `wikidata`, `google-kp`, `wikipedia`, or `directory`; `--status` takes `present`, `absent`, `inconsistent`, or `outdated`):

   python "${CLAUDE_PLUGIN_ROOT}/scripts/geo-tracker.py" \
       --brand {slug} --action entity-check \
       --platform wikidata \
       --entity-name "Acme Corp" \
       --status inconsistent \
       --details "Foun