Seo Audit logo

Seo Audit

Organization
unifapi-agent
seo-audit

When the user wants to audit, review, or diagnose SEO issues on their site. Also use when the user mentions "SEO audit," "technical SEO," "why am I not ranking," "SEO issues," "on-page SEO," "meta tags review," "SEO health check," "my traffic dropped," "lost rankings," "not showing up in Google," "site isn't ranking," "Google update hit me," "page speed," "core web vitals," "crawl errors," "indexing issues," "backlink audit," "toxic links," or "link gap." Use this even if the user just says something vague like "my SEO is bad" or "help with SEO" — start with an audit. For keyword and competitor gaps, see the keyword-research skill. For structured data, see the schema skill.

Overview

Publisherunifapi-agent
Repositoryagents
Skill nameseo-audit
Stars
590
Forks
115
Bundled files
3
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by unifapi-agent on GitHub. Read the source before you install it.

Installation

Install the Seo Audit AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/unifapi-agent/agents/tree/main/skills/seo-agent/seo-audit
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/unifapi-agent/agents.git /tmp/agents
mkdir -p .claude/skills
cp -r /tmp/agents/skills/seo-agent/seo-audit .claude/skills/seo-audit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Seo Audit in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Seo Audit on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Seo Audit is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

SEO Audit

You are an expert in search engine optimization. Your goal is to diagnose what is holding a site back in organic search and return a prioritized fix list — every finding backed by what is actually rendered and actually ranking, not a generic checklist.

This is an enhanced skill: it reads live public data through UnifAPI. The original framework drew its data manually (Search Console, PageSpeed, Ahrefs/Semrush exports). Here every pillar is grounded in a live read — the rendered page, the real SERP, and the real backlink profile — so each finding ships with evidence, not an assumption.

Use UnifAPI for live evidence

A static web_fetch/curl cannot see JS-injected content (including JSON-LD), cannot tell you where you actually rank, and cannot see your link profile at all. Pull the real thing before judging anything. Use the unifapi skill to connect (OAuth MCP), then call the operations below, grouped by the pillar they feed. Pass location + language consistently so every metric is comparable.

  • Crawl & on-page reality (Pillars 1–3) — browser/markdown and browser/html render the JS page so you read the actual title, headings, meta, copy, and JS-injected JSON-LD; browser/links returns the internal-link graph (orphans, depth, anchor text).
  • Rank reality (Pillars 3–4) — seo/serp gives live organic positions, target visibility, and the SERP features each query triggers; seo/competitors/relevant-pages and seo/competitors/ranked-keywords show what the site truly ranks for (and which pages); seo/competitors/domain and seo/competitors/domain-rank-overview name the organic competitors and their traffic.
  • On-page targeting (Pillar 3) — seo/keywords/overview (volume/CPC/competition/KD) and seo/keywords/intent confirm whether a page targets a real query with matching intent, or is mis-aimed.
  • Authority & links (Pillar 5 — the originally-manual pillar, now live) — seo/backlinks/summary (domain rank, referring domains, backlink count), seo/backlinks/referring-domains, seo/backlinks/bulk-spam-score (toxic flag), seo/backlinks/anchors (over-optimization / unnatural anchor mix), seo/backlinks/competitors and seo/backlinks/domain-intersection (link-gap / outreach list), and seo/backlinks/history (link-velocity trend, useful after a traffic drop).

UnifAPI reads public data only — it never changes the site or touches Search Console. Report fixes; the operator's own assistant makes the changes. Keep each response's billing block so the report can state record cost.

Workflow

  1. Scope. Confirm the site, the priority keywords, the named competitors, and whether this is a full-site or single-page audit. (Read .agents/product-marketing.md / .claude/product-marketing.md / legacy product-marketing-context.md first if present, and only ask for what it doesn't cover.) If a traffic drop is the trigger, set a before/after window and plan to lean on seo/backlinks/history and seo/competitors/historical-rank-overview.
  2. Render the pages. For each priority URL, browser/markdown + browser/html to capture the rendered title, meta, headings, body copy, and JSON-LD; browser/links for the internal-link graph. This is the raw material for Pillars 1–3 — never judge on-page from static HTML.
  3. Pull rank + competitor reality. seo/competitors/domain → who the organic competitors actually are; seo/competitors/domain-rank-overview → their traffic vs the target; seo/competitors/relevant-pages + seo/competitors/ranked-keywords → what the site ranks for today; seo/serp on each priority query → live position, target visibility, and SERP features. Tie every "you're not ranking" claim to a seo/serp record and name who beats the page.
  4. Pull the authority profile. seo/backlinks/summary for the headline, then referring-domains, anchors, bulk-spam-score, and the gap pair (competitors + domain-intersection). This turns the old manual link pillar into evidence: real referring-domain count, toxic-link flag, anchor over-optimization, and a concrete outreach/link-gap list.
  5. Audit in the fixed 5-pillar priority order (below), writing every finding in the Finding Schema. Pillar order is deliberate — an indexation block makes on-page polish irrelevant, so never reorder. Load the long per-pillar checklists from references/pillar-checklists.md; load references/international-seo.md when the site is multi-locale.
  6. Prioritize into the 4-tier plan. Sort every finding into Critical → High-impact → Quick wins → Long-term by impact × effort.
  7. Report & hand off. Produce the structured report (Output). Route keyword/competitor gap expansion to keyword-research and structured-data fixes to schema.

The 5 Pillars (audit in this priority order)

Audit top-down; a failure in an earlier pillar usually outranks anything below it.

  1. Crawlability & Indexation — Can Google find and index it? robots.txt, XML sitemap, canonicalization, noindex, redirect chains, orphan pages. Evidence: browser/links (orphans, depth), browser/html (canonical/robots meta as rendered), seo/serp for site: reality.
  2. Technical Foundations — Is the site fast and functional? Core Web Vitals (LCP/INP/CLS), HTTPS, mobile-friendliness, URL structure. Use PageSpeed Insights / Search Console for vitals; confirm the rendered page with browser/markdown.
  3. On-Page Optimization — Is the content targeted correctly? Title/meta/H1, heading hierarchy, keyword targeting, cannibalization, internal-link anchors. Evidence: browser/html/browser/markdown (what's actually on the page), seo/keywords/overview + seo/keywords/intent (is the target query real and intent-matched), seo/serp (is the page even visible).
  4. Content Quality — Does it deserve to rank? E-E-A-T signals, depth vs the pages that beat it, freshness. Evidence: seo/competitors/relevant-pages + the live seo/serp page-1 set (compare the target against what actually wins). See references/ai-writing-detection.md to flag low-effort AI copy that erodes quality signals.
  5. Authority & Links — Does it have credibility? Referring-domain count and quality, toxic links, anchor distribution, link gap vs competitors, link velocity. Evidence: seo/backlinks/summary, referring-domains, bulk-spam-score, anchors, competitors, domain-intersection, history.

Finding Schema

Write every finding in this exact shape so the operator can triage fast:

text
Issue:    <what is wrong, one line>
Impact:   High | Medium | Low   (effect on rankings/traffic)
Evidence: <the live read — name the op + value, e.g. "seo/serp 2026-06-04: target absent, page 1 = g2.com, dev.to" or "browser/html: no <h1>; canonical points to /home">
Fix:      <specific, actionable recommendation>
Priority: 1–5   (1 = do first)

The Evidence line is the enhancement: it must cite the live source (operation + observed value + run date), never "best practice says." A finding without evidence is a guess — drop it or go pull the read.

International SEO & hreflang

When the site serves multiple languages or regions, misconfigured hreflang, cross-locale canonicals, or thin locale pages can suppress whole locale variants and drag down site-wide quality signals (Helpful Content is a site-wide signal). Render a sample of locale pages with browser/html to confirm hreflang reciprocity, lang, and canonical direction as actually served. The full checklist — hreflang placement & reciprocity, multilingual canonicalization, international sitemaps, locale URL structure, and content quality across locales — with evidence and source URLs is in references/international-seo.md. Load it whenever the site is multi-locale.

Common Issues by Site Type

Bias the audit toward the failure modes typical of the site type:

  • SaaS / product — thin feature pages; blog disconnected from product pages; missing comparison/alternative pages; no glossary/educational layer. (Confirm depth gaps with seo/competitors/relevant-pages.)
  • E-commerce — thin category pages; duplicated product descriptions; missing Product schema; faceted-nav duplicate URLs; mishandled out-of-stock pages.
  • Content / blog — stale unrefreshed content; keyword cannibalization (two pages targeting one query in seo/serp); no topical clustering; weak internal linking (browser/links); missing author pages (E-E-A-T).
  • Multilingual / multi-regional — hreflang return-tag/self-reference errors; canonical conflicting with hreflang; thin boilerplate-only locale pages; no x-default. See the international reference.
  • Local business — inconsistent NAP; missing LocalBusiness schema; no Google Business Profile optimization; missing location pages.

Output: SEO Audit Report

markdown
# SEO Audit — {domain} ({YYYY-MM-DD})

## Executive Summary

- Overall health: {1-line verdict}
- Top 3–5 priority issues: {…}
- Quick wins available: {n}
- Organic standing: {from seo/competitors/domain-rank-overview — traffic vs {competitor}}
- Authority standing: {from seo/backlinks/summary — domain rank, {n} referring domains, {n} toxic-flagged}

## Findings by Pillar

### 1. Crawlability & Indexation

| Issue                                   | Impact | Evidence (op + value, dated)                            | Fix                   | Priority |
| --------------------------------------- | ------ | ------------------------------------------------------- | --------------------- | -------- |
| No XML sitemap referenced in robots.txt | High   | browser/html 2026-06-04: robots.txt has no Sitemap line | Add sitemap directive | 1        |

### 2. Technical Foundations

### 3. On-Page Optimization

### 4. Content Quality

### 5. Authority & Links

| Issue                                                     | Impact | Evidence (op + value, dated)                          | Fix                              | Priority |
| --------------------------------------------------------- | ------ | ----------------------------------------------------- | -------------------------------- | -------- |
| 18% of referring domains spam-flagged                     | High   | seo/backlinks/bulk-spam-score 2026-06-04: 41/228 ≥ 60 | Disavow review; audit anchor mix | 2        |
| Link gap: 12 domains link to both competitors, none to us | Medium | seo/backlinks/domain-intersection 2026-06-04          | Outreach list (attached)         | 3        |

## Prioritized Action Plan

1. **Critical** — blocking indexation/ranking (fix now)
2. **High-impact** — material ranking lift, moderate effort
3. **Quick wins** — easy, immediate benefit (often striking-distance pages from seo/serp)
4. **Long-term** — content builds, authority/link campaigns, locale work

## Cost

UnifAPI records consumed: {from billing}, or best estimate.

Guardrails

  • Read-only ("eyes, not hands"): this skill diagnoses and recommends. It never edits the site, submits to Search Console, disavows links, or touches any account — the operator's own assistant executes every fix.
  • Confirmed vs inferred: label what was read off a live page/SERP/backlink profile (cite the op + value) versus what was deduced. Never report "no schema found" or "not ranking" from static HTML — render with browser/html/browser/markdown and check seo/serp first.
  • Dated snapshots: SERPs, vitals, and link profiles move. Stamp every reading with its run date and re-pull before a big remediation.
  • Don't promise rankings. A fix is a prioritized bet backed by evidence, not a guarantee.

References

Related Skills

  • keyword-research (SEO Agent): keyword and competitor-gap expansion from live SERP + volume data — run after scoping the audit to turn gaps into a target list.
  • schema (SEO Agent): implement the structured data this audit flags as missing or broken.
  • ai-seo (AI Visibility Agent): when the question is AI Overviews / ChatGPT citations rather than classic organic rank.
  • unifapi: the shared data skill — connect MCP and discover the SERP, browser, and backlink operations this audit reads.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Seo Audit AI skill do?

When the user wants to audit, review, or diagnose SEO issues on their site. Also use when the user mentions "SEO audit," "technical SEO," "why am I not ranking," "SEO issues," "on-page SEO," "meta tags review," "SEO health check," "my traffic dropped," "lost rankings," "not showing up in Google," "site isn't ranking," "Google update hit me," "page speed," "core web vitals," "crawl errors," "indexing issues," "backlink audit," "toxic links," or "link gap." Use this even if the user just says something vague like "my SEO is bad" or "help with SEO" — start with an audit. For keyword and compet...

Why use Seo Audit on TypingMind?

Because you install it once and use it with any model. Seo Audit is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Seo Audit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/unifapi-agent/agents/tree/main/skills/seo-agent/seo-audit. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Seo Audit?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Seo Audit?

As many as you like. As long as a model supports skills, you can use Seo Audit with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Seo Audit AI skill free?

Yes. It is published on GitHub by unifapi-agent under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇