agentleFS
martinbattaglia-commitsmartinbattaglia-commits/Claude-skills/AGENTS.md

Claude-skills

AGENTS.md0 starsChanged 5 months agoMITView on GitHub

For Cursor, Cursor Cloud Agents, Google Antigravity, Gemini CLI, OpenAI Codex CLI, Cline, Aider, and any other agent harness that reads project-root agent instructions. Claude Code users: see CLAUDE.md instead. Every skill in skills/*/SKILL.md is authored to a portable subset of the Claude Code skill spec. Validate compatibility with your harness via: The check confirms each SKILL.md has the minimum frontmatter every harness expects (name, description, optional model, optional tools) and warns on Claude-Code-specific features (maxTurns, multi-line tool list with…

The file

We split this AGENTS.md into the rules it gives an agent

4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Every skill in skills/*/SKILL.md is authored to a portable subset of the Claude Code skill spec.

    line 11
  2. Very specific:

    Place the repo in ~/.gemini/antigravity/plugins/claude-seo/ or install via:

    line 123
  3. Very specific:

    Progressive Disclosure: Read SKILL.md for routing, load references on demand

    line 164
  4. Very specific:

    Security: All scripts call validate_url() for SSRF protection

    line 166

≈ “Common” is an estimate: at least 1 in 5 of the 41,581 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Claude SEO: Multi-Platform Agent Instructions

For Cursor, Cursor Cloud Agents, Google Antigravity, Gemini CLI, OpenAI Codex CLI, Cline, Aider, and any other agent harness that reads project-root agent instructions.

Claude Code users: see CLAUDE.md instead.

Cross-platform portability (v2.0.0)

Every skill in skills/*/SKILL.md is authored to a portable subset of the Claude Code skill spec. Validate compatibility with your harness via:

python3 scripts/portability_check.py

The check confirms each SKILL.md has the minimum frontmatter every harness expects (name, description, optional model, optional tools) and warns on Claude-Code-specific features (maxTurns, multi-line tool list with descriptive comments) that other harnesses may ignore but do not reject.

Per-harness notes

HarnessHow to load claude-seo
CursorSymlink or copy skills/ and agents/ into .cursor/rules/. Commands are invoked as text prompts; the harness reads SKILL.md body as system context.
Cursor Cloud AgentsPush the repo; Cloud Agents read AGENTS.md automatically at session start.
Google AntigravityPoint the workspace at this repo root; Antigravity reads AGENTS.md first, falls back to skills/.
Gemini CLIgemini init in this repo loads AGENTS.md. Skills are activated via activate_skill <name> in conversation.
OpenAI Codex CLIReads AGENTS.md from project root. Bash tools work as documented; some Claude-specific tool names (Read/Write/Edit) are aliased to Codex equivalents transparently.
ClineLoads AGENTS.md from project root. Skills appear as system messages; subagent delegation falls back to in-context expansion.
AiderReads AGENTS.md if present; otherwise falls back to README. Aider does not support sub-agent dispatch; the seo-* skills run inline.

Tool-name compatibility

Where claude-seo skills mention Claude Code tools (Read, Write, Edit, Bash, Glob, Grep, WebFetch), each harness typically has an equivalent:

Claude CodeCodexClineAiderCursor / Antigravity
Readread_fileread_file(inline)read
Writewrite_filewrite_file/add then editwrite
Editapply_diffreplace_in_file/editedit
Bashbashexecute_command/runshell
Globglobsearch_files(inline)find
Grepgrepsearch_files/grepgrep
WebFetchfetch / browse(browser tool)(n/a)fetch

These mappings are automatic in most harnesses; we list them for transparency in case a recipe needs a specific call.

Overview

Claude SEO is a Tier 4 SEO analysis skill with 26 sub-skills (21 core + 1 orchestrator + 1 framework integration + 2 extension mirrors + 1 security), 18 sub-agents (15 core + 1 framework integration + 2 extension mirrors), and 50 Python execution scripts.

Quick Reference

CommandWhat it does
/seo audit <url>Full website audit with parallel subagent delegation
/seo page <url>Deep single-page analysis
/seo technical <url>Technical SEO audit (9 categories)
/seo content <url>E-E-A-T and content quality analysis
/seo schema <url>Schema.org detection, validation, generation
/seo sitemap <url>XML sitemap analysis or generation
/seo images <url>Image SEO: on-page audit, SERP analysis, file optimization
/seo geo <url>AI Overviews / Generative Engine Optimization
/seo plan <type>Strategic SEO planning
/seo cluster <keyword>SERP-based semantic clustering and content architecture
/seo sxo <url>Search Experience Optimization: page-type analysis, personas
/seo drift baseline <url>Capture SEO baseline for change monitoring
/seo drift compare <url>Compare current state to stored baseline
/seo drift history <url>Show drift history over time
/seo ecommerce <url>E-commerce SEO: product schema, marketplace intelligence
/seo programmatic [url]Programmatic SEO at scale
/seo competitor-pages [url]Competitor comparison pages
/seo local <url>Local SEO analysis (GBP, citations, reviews)
/seo maps [cmd] [args]Maps intelligence (geo-grid, GBP audit, competitors)
/seo hreflang <url>Hreflang/i18n SEO audit, cultural profiles, content parity
/seo google [cmd] [url]Google SEO APIs (GSC, PageSpeed, CrUX, Indexing, GA4)
/seo backlinks <url>Backlink profile analysis
/seo backlinks setupSetup free backlink APIs
/seo backlinks verify <url>Verify known backlinks still exist
/seo dataforseo [cmd]Live SEO data via DataForSEO (extension)
/seo image-gen [use-case]AI image generation for SEO assets (extension)
/seo firecrawl [cmd] <url>Full-site crawling and site mapping (extension)

Using with Cursor / Cursor Cloud

Cursor reads this file automatically. All SKILL.md files contain the full analysis logic as natural language instructions. Python scripts in scripts/ provide execution capabilities.

Running scripts directly (Cursor doesn't have MCP):

# Page fetching with SSRF protection
python3 scripts/fetch_page.py https://example.com

# HTML parsing for SEO elements
python3 scripts/parse_html.py https://example.com

# PageSpeed Insights
python3 scripts/pagespeed_check.py https://example.com --json

# Drift baseline
python3 scripts/drift_baseline.py https://example.com

# DataForSEO (requires credentials)
DATAFORSEO_USERNAME=user DATAFORSEO_PASSWORD=pass python3 scripts/dataforseo_merchant.py search "keyword"

Cursor Cloud gotchas:

  • SSL certificates may not resolve for some domains. Investigate the certificate issue rather than disabling verification.
  • PATH may not include Python venv. Use full path: ~/.claude/skills/seo/.venv/bin/python
  • Screenshots save to /tmp/ not CWD. Check absolute paths.

Using with Google Antigravity

Antigravity discovers this project via .claude-plugin/plugin.json. Place the repo in ~/.gemini/antigravity/plugins/claude-seo/ or install via:

bash install.sh

Architecture

skills/                    # 26 sub-skills (auto-discovered)
  seo/SKILL.md            # Main orchestrator + routing
  seo-cluster/            # Semantic clustering (v1.9.0)
  seo-sxo/                # Search Experience Optimization (v1.9.0)
  seo-drift/              # SEO drift monitoring (v1.9.0)
  seo-ecommerce/          # E-commerce SEO (v1.9.0)
  seo-audit/              # Full site audit
  seo-page/               # Single-page analysis
  seo-technical/          # Technical SEO
  seo-content/            # E-E-A-T quality
  seo-schema/             # Schema.org markup
  seo-sitemap/            # XML sitemaps
  seo-images/             # Image optimization
  seo-geo/                # AI search / GEO
  seo-local/              # Local SEO
  seo-maps/               # Maps intelligence
  seo-plan/               # Strategic planning
  seo-hreflang/           # International SEO
  seo-google/             # Google APIs
  seo-backlinks/          # Backlink analysis
  seo-programmatic/       # Programmatic SEO
  seo-competitor-pages/   # Competitor pages
  seo-dataforseo/         # DataForSEO (extension)
  seo-image-gen/          # AI images (extension)
agents/                    # 18 subagents
scripts/                   # 50 Python scripts
schema/                    # JSON-LD templates
extensions/                # Optional add-ons (DataForSEO, Firecrawl, Banana, ASO)

Key Principles

  1. Progressive Disclosure: Read SKILL.md for routing, load references on demand
  2. Industry Detection: Auto-detect SaaS, e-commerce, local, publisher, agency
  3. Security: All scripts call validate_url() for SSRF protection
  4. Config location: ~/.config/claude-seo/ for API credentials

Credits

Created by @AgriciDaniel. v1.9.0 community contributions by Lutfiya Miller, Chris Muller, Florian Schmitz, Dan Colta, and Matej Marjanovic. See CONTRIBUTORS.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.