15 added, 12 removed. Audit A to A.
---
name: audit
- description: Audit a website or web codebase for SEO and AI-search (GEO/AEO) — produces two independent 0-100 scores (Search SEO + AI Visibility) plus a prioritized, evidence-backed report. Read-only; never writes files. Use when the user asks to audit, analyze, check, or score a site's SEO, structured data/schema, meta tags, Core Web Vitals, robots.txt, sitemaps, or llms.txt, or how it ranks in Google and is cited by AI engines (ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude).
- argument-hint: "<url|path> [--render auto|static|js] [--vertical auto|saas|blog-publisher|local-business|ecommerce|docs]"
- allowed-tools: Read, Grep, Glob, WebFetch, Bash, Task
+ description: Audit a website or web codebase for SEO and AI-search (GEO/AEO) — produces two independent 0-100 scores (Search SEO + AI Visibility) plus a prioritized, evidence-backed report persisted on disk. Read-only; never writes to the project. Use when the user asks to audit, analyze, check, or score a site's SEO, structured data/schema, meta tags, Core Web Vitals, robots.txt, sitemaps, or AI discovery files, or how it ranks in Google and is cited by AI engines (ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude).
+ argument-hint: "<url|path> [--pages N] [--max N] [--render auto|static|js] [--ua default|googlebot|gptbot] [--vertical auto|saas|blog-publisher|local-business|ecommerce|docs] [--out <dir>]"
+ allowed-tools: Read, Grep, Glob, WebFetch, Bash, Agent
---
# /claude-seo-ai:audit
- A full, **read-only** SEO + AI-search audit. Never writes files.
+ A full, **read-only** SEO + AI-search audit. Never writes to the user's project; artifacts go to `${CLAUDE_PLUGIN_DATA}/runs` (or `--out`).
- `$ARGUMENTS` = `<url|path> [flags]`. The target is a URL (`https://…`) or a local web project / built HTML. If no target is given, ask for one.
+ `$ARGUMENTS` = `<url|path> [flags]`. The target is a URL (`https://…`) or a local web project / built HTML. If no target is given, ask for one. Flags (`--pages`, `--max`, `--render`, `--ua`, `--vertical`, `--out`) are forwarded verbatim to the orchestrator. <!-- reconcile-flags-v0.2 -->
## What to do
- 1. Invoke the **seo-orchestrator** skill with the target and any flags. It builds the shared PageSnapshot (via **seo-crawl-render**), detects the vertical (**seo-vertical-detect**, see `references/routing.md`), dispatches the read-only specialist subagents in parallel, merges findings, and runs **seo-score**.
- 2. Present:
- - The two scores — **Search SEO** and **AI Visibility** — each with a band (A–F) and a one-line interpretation.
- - A per-category breakdown for each score, the data **tier** reached, and the count of `needs_api` checks (score confidence).
- - A **prioritized action list** sorted by impact ÷ effort: each item with status, evidence, recommendation, fixability (auto/proposed/advisory), and `expected_impact` (axis + confidence + magnitude).
- 3. End by offering: "Run `/claude-seo-ai:fix` to apply the safe, deterministic fixes — you'll confirm each change."
+ 1. Invoke the **seo-orchestrator** skill with the target and all flags. It acquires the site with the bundled scripts (crawl/snapshot → PageSnapshot v2), detects platform + vertical (**seo-vertical-detect**, `references/routing.md`), runs the deterministic checks, dispatches the read-only specialist subagents in parallel with a dispatch envelope, merges everything with `report.mjs`, and scores.
+ 2. Present, from `report.json`:
+ - The two scores — **Search SEO** and **AI Visibility** — each with a band (A–F) and a one-line interpretation. Never blended.
+ - A per-category breakdown for each score; **coverage mode** (full vs deterministic-only), the data **tier** reached, and the `needs_api` / `manual_review` counts (score confidence).
+ - The **sampling table**: pages crawled by role, templates discovered vs sampled, and every skip with its reason.
+ - The **platform profile** (platform/framework/plugins, environment — e.g. "noindex expected on preview").
+ - A **prioritized action list** sorted by impact ÷ effort: status, evidence, recommendation, fixability (auto/proposed/advisory), `expected_impact` (axis + confidence + magnitude).
+ - The absolute **report path** (`<run_dir>/report.md`).
+ 3. End by offering: "Run `/claude-seo-ai:fix <target>` to apply the safe, deterministic fixes — you'll confirm each change" and "`/claude-seo-ai:compare --baseline latest --against <target>` after you deploy".
- Two scores, never blended. Every finding conforms to `schema/finding.schema.json` with reproducible evidence. Respond in the user's language (EN/ES).
+ Every finding conforms to `schema/finding.schema.json` with reproducible evidence. Respond in the user's language (EN/ES).