The open-source SEO skills for coding agents.
Drop 36 repeatable SKILL.md workflows into Claude Code, Cursor, Codex, and OpenCode. Stop writing AI slop and start shipping pages that rank.
Flagship skills(6)
SEO skill
/blitz-seo
Runs a focused 30-day SEO sprint that concentrates search-intent alignment, internal link equity, first-hand proof, and targeted backlinks onto a single money page to push it into the Top 3. Use when a high-value page is stuck on Page 2, the user mentions "blitz seo" or a 30-day SEO sprint, or SEO effort is spread too thin across dozens of URLs.
# Blitz SEO: The 30-Day Sprint to Rank One PageGuides an aggressive, short-term SEO sprint that concentrates every available resource on ranking a single high-value page for one money keyword (1-3 max). Not scale, not content volume: focus, speed, and pressure.**When invoking:** On **first use**, if helpful, open with 1-2 sentences on what this skill covers and why it matters, then provide the main output. On **subsequent use** or when the user asks to skip, go directly to the main output.## Core TruthMost SEO fails not because it is bad, but because it is spread too thin. You do not need hundreds of pages ranking past position 10. You need 1-3 money pages in the top 3.## Initial Assessment**Check for project context first:** if `.cursor/project-context.md` or `.claude/project-context.md` exists, read it for product, audience, and current rankings. Otherwise identify:1. **Money keyword candidates**: high-intent, high-volume keywords tied to the core offering. Not blog keywords, not long-tails.2. **Current position**: a page stuck at position 5-20 is the ideal Blitz target (already indexed, needs pressure).3. **Revenue link**: would ranking top 3 for this keyword directly produce signups or sales?4. **Available resources**: backlink budget, existing blog posts for internal links, social accounts for signals.**The target test:** your Blitz keyword should be something that _hurts to lose_ to competitors.## The Sprint FrameworkSEO skill
/seo-roast
Runs a brutal, revenue-first SEO teardown of a site and its codebase, kills fatal crawl, programmatic, layout, authority, and AI-visibility blockers in a fix PR, and ships a revenue-ordered seo/roast-roadmap.md (plus an optional client deck). Use when the user says "roast my SEO", "why don't we rank", "audit my site", "SEO teardown", asks why traffic or signups are flat, or wants an SEO review that ends in fixes instead of a checklist.
# SEO Roast: Revenue-First Teardown That Ends in a Merged PRMost audits ship a 50-page PDF of 400 warnings: missing alt text, meta descriptions 4 characters too long, H3 before H2. The site keeps losing money because the 3 issues that matter sit on page 31. A roast does the opposite: find what blocks revenue, say it bluntly, fix it in code, and order the rest by money.**When invoking:** Open with a one-line verdict ("Your robots.txt hides you from every engine except Google"), then run the procedure. Never open with a score out of 100.## Core Rule: Fatal First, Fluff Never (Until Fatal Is Dead)```issue found├── FATAL: blocks crawling, indexing, or rendering of a revenue page│ └── Fix in this PR. Nothing else ships until this tier is empty.├── REVENUE: page indexes but cannot rank or cannot convert│ └── Fix in this PR if it takes under 1 hour of code, else top of roadmap.├── COMPOUNDING: authority, links, AI visibility│ └── Roadmap with owner and first action.└── FLUFF: alt text, meta length, heading order, minor schema warnings └── One line at the bottom of the roadmap. Do not fix while FATAL or REVENUE rows remain.```Why: an alt tag on a page Bing cannot crawl earns nothing, and teams that fix fluff first feel productive while the blocker stays live for another quarter.Content skill
/unslop-seo
Run an in-place editorial de-slopping and structural Answer-Engine (GEO / LLM citation) HTML pass across MDX blog posts, landing pages, and documentation. Use when cleaning up AI-generated drafts, grounding vague marketing copy in real codebase facts, optimizing headings and opening paragraphs for ChatGPT and Perplexity citations, or removing AI tells like em dashes and puffery.
# Unslop SEO and answer-engine passAI-edited posts fail twice: engineers recognize the filler and leave, and the direct answer sits under 200 words of preamble where answer engines never quote it. This skill edits MDX, Markdown, TSX, or HTML files **in place** in three phases, in order: ground the facts, fix the structure, then strip the slop. Facts come first because rewriting prose that later gets deleted wastes the pass.## Pick the target files```Did the user name files or URLs?├── Yes -> edit exactly those└── No -> run `git diff --name-only HEAD~5 -- '*.md' '*.mdx' '*.tsx' '*.html'` ├── Content files changed -> edit those └── None -> list `content/blog/`, `app/**/page.tsx`, `docs/` and edit the 3 newest posts```## Decision tree per file```Phase 1: Does every product claim name a mechanism or a number?├── Check `seo/kb/brand-ssot.md`, `/pricing`, `/about`, or source code├── Feeling words ("blazing-fast", "seamless") -> exact mechanics ("compiles 1,200 MDX routes in 1.4s")└── Sentence could appear unchanged on a competitor's site -> delete or rewritePhase 2: Is the page built for featured snippets and answer-engine chunking?SEO skill
/ahrefs-research
Pulls Ahrefs data through the official MCP server or API v3 and ships a competitor gap Kanban with scaffolded pages, a clean link-replication CSV, Reddit-mined page specs, and exact link-insertion target URLs. Use when the user asks for competitor research, content gaps, top pages, backlink audits, 'is this DR real', Reddit topic mining, or picking a page on a prospect site for a link insertion.
# Ahrefs Research for Coding AgentsAhrefs data is only worth the API units if it ends as a file in the repo. Every workflow below terminates in a Markdown roadmap, a CSV, or a scaffolded route. Never stop at "here is what I found".**When invoking:** On first use, state which connection path you detected and which workflow you are running, then run it. On repeat use, go straight to the output.## Step 0: Pick the Connection Path```Is an Ahrefs tool in your tool list (names mirror v3 paths, e.g. site-explorer-top-pages)?├── Yes → use the MCP tools. Same params as the REST calls below.└── No → is $AHREFS_API_KEY set? (run: test -n "$AHREFS_API_KEY" && echo set) ├── Yes → use the `ahrefs` shell helper below (REST v3). └── No → CSV fallback: 1. Ask the user to export the matching Ahrefs report as CSV (UTF-8). 2. Save to seo/data/ahrefs/<report>-<domain>.csv 3. Parse the CSV columns (URL, Traffic, Top keyword, Value, Referring domains) and continue the same workflow from the "Filter" step.```Why the order: MCP needs zero glue code, REST is scriptable and repeatable, CSV is slow but keeps the workflow alive when the user has a UI seat and no API plan.SEO skill
/seo-title-optimization
Rewrite and audit SEO page titles, title tags, MDX frontmatter titles, H1 alignment, and bulk title improvements. Use when the user mentions titles, title tags, SERP titles, metadata titles, SEO titles, CTR, truncation, or asks to improve titles across pages.
# SEO Title OptimizationUse this skill to improve page titles for search rankings and click-through rate.## Core RuleTreat the SERP as the brief.Do not write titles from vibes. Identify the target query, study what already ranks when possible, then write the clearest title that matches the page and search intent.## Title FormulaStart with:```textPrimary keyword + modifier / benefit + brand```Examples:- `Client Portal Software for Accountants`- `Best AI Presentation Makers in 2026`Code skill
/eval
Build, run, or update minimal evaluation benchmarks for AI prompts, LLM calls, multi-model comparisons, and core functions. Use when creating new eval scripts, comparing models side-by-side, filtering specific candidate models, running benchmarks, or inspecting eval.md output artifacts.
# Eval SystemA clean, minimal benchmark system that runs production code against test cases and writes results to a co-located markdown artifact (`eval.md`). The `eval.md` is a **judgment surface**: it exists so a strong external model or a human can read the real outputs and judge quality, not so the eval can auto-declare a winner.---## Core Principles & Rules### 1. Import Production Logic Directly- **ALWAYS** import and invoke the actual production function, prompt constants, and schemas from the feature file.- **NEVER** duplicate prompts, system messages, or JSON schemas inside `eval.ts`.- *Why:* Re-implementing prompts in the eval causes prompt drift. When production changes, a duplicated eval becomes a false benchmark testing obsolete behavior.### 2. Benchmark Real Runs — Never Estimate- **ALWAYS** execute every candidate model to capture ground-truth latency, tokens, and cost.- **NEVER** estimate or hallucinate latency, token counts, or pricing in a report without running the models.- *Why:* Models cannot predict real streaming latency, throughput, or cost. Only direct execution records the truth.### 3. "Passed" (✅) Means Schema Valid — NOT High Quality- **ALWAYS** treat the status column as execution correctness (did the call succeed and conform to the Zod schema?).- **NEVER** assume a "Passed" status implies the output is accurate, relevant, or good.- *Why:* A model can return 100% valid JSON conforming to a schema while outputting completely wrong or noisy content.All 36 skills
- SEO (20)
- /ahrefs-research
- /ai-seo-report
- /aiseotracker
- /awesome-tools
- /blitz-seo
- /block-ai-scrapers
- /blog-refresher
- /content-coverage
- /dataforseo-serp
- /directory-submissions
- /domain-migration
- /honest-link-outreach
- /internal-linking
- /linkdr
- /proof-kb
- /reddit-serp
- /seo-autopilot
- /seo-roast
- /seo-title-optimization
- /vs-and-alternatives
- Copywriting (2)
- Marketing (3)
- Product (2)
Guide: how the skills chain togetherFree 71-point SEO launch checklist
FAQ
- What is an agent skill?
- A folder with a SKILL.md Markdown file that teaches an agent such as Claude Code, Cursor or Codex one repeatable workflow, like clustering live SERPs, rewriting title tags or planning backlinks.
- How do I install them in Claude Code and Cursor?
- Run npx skills add Illyism/seo-skills in your project or home directory. Claude Code and Cursor read each skill's name and description and load the full file when your prompt matches.
- How do skills work with MCP servers?
- The SKILL.md gives the agent the method and the approval steps. An MCP server gives it live data, such as Search Console queries, SERPs or AI citations. Use both together.