Seo Agent logo

Seo Agent

Community
TheMattBerman
seo-agent

Self-improving SEO system. Finds keywords via DataForSEO, writes content that ranks, monitors GSC for what's climbing, then writes MORE content to push winners higher. A feedback loop, not a one-shot tool.

Overview

PublisherTheMattBerman
Repositoryseo-kit
Skill nameseo-agent
Stars
100
Forks
26
Bundled files
3
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 TheMattBerman on GitHub. Read the source before you install it.

Installation

Install the Seo Agent 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/TheMattBerman/seo-kit/tree/main/skills/seo-agent
  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/TheMattBerman/seo-kit.git /tmp/seo-kit
mkdir -p .claude/skills
cp -r /tmp/seo-kit/skills/seo-agent .claude/skills/seo-agent
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Seo Agent 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 Agent 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 Agent 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 Agent — The Self-Improving SEO Loop

Most SEO tools do one thing. Discover keywords. Write content. Check rankings. Done.

This isn't that.

The SEO Agent is a feedback loop — a system that finds opportunities, acts on them, watches what happens, and feeds those results back into the next round of work. Every cycle makes the next cycle smarter. Rankings climb. The content library grows. And you're always pushing the keywords that are already moving instead of starting from scratch.

"Don't just rank. Rank, learn, and rank higher."


The Loop

DISCOVER → WRITE → MONITOR → OPTIMIZE → REPEAT

1. DISCOVER (seo-discover.sh) Pull your current GSC rankings. Find what's in the strike zone (positions 5-20). Hit DataForSEO for related keywords you're missing. Score opportunities by volume, competition, and current position. Output a prioritized hit list.

2. WRITE (You + Claude) Take the top opportunity from discovery. Generate publication-ready SEO content — outline, draft, optimized headers, FAQs, schema markup. Brand voice from workspace/brand/voice-profile.md baked in.

3. MONITOR (seo-monitor.sh) Run weekly. Pull current GSC data. Compare to last snapshot. Flag what's climbing, what's dropping, what just cracked the strike zone. This is your radar.

4. OPTIMIZE (seo-compete.sh + Claude) Take the winners from monitoring — content climbing from 15 → 8. Write supporting articles that link to them. Add internal links from existing content. Push the climbers over the finish line.

5. REPEAT Discovery now has more context. Monitor has more history. Every loop compounds.


The Strike Zone

Positions 5–20 are where SEO leverage lives.

  • Position 1–4: Already winning. Defend.
  • Position 5–20: ⚡ THE STRIKE ZONE — close enough to move, not yet getting the clicks. A supporting article, better internal links, or a refreshed page can jump these 5–10 spots. That's the difference between 2% CTR and 8% CTR.
  • Position 21+: Too far out. Needs a full content build.

The Agent focuses energy here by default.


Scripts

All scripts live in ~/clawd/skills/seo-agent/scripts/. All use bash + curl + jq. No Python. No npm.

seo-discover.sh — Find Your Next Opportunity

bash
./seo-discover.sh --site sc-domain:example.com [--seeds "keyword1,keyword2"] [--limit 20]

What it does:

  1. Pulls your GSC top queries (last 28 days)
  2. Filters for strike zone keywords (positions 5–20)
  3. Hits DataForSEO for keyword suggestions around your top queries
  4. Gets search volume + competition for all candidates
  5. Scores and ranks opportunities
  6. Outputs a prioritized JSON list

Output fields: keyword, current_position, search_volume, competition, opportunity_score

Add --json for raw machine-readable output. Default is a human-readable table.

seo-monitor.sh — Track What's Moving

bash
./seo-monitor.sh --site sc-domain:example.com [--days 28]

What it does:

  1. Pulls current GSC data
  2. Loads previous snapshot from ~/clawd/workspace/seo-agent/snapshots/
  3. Diffs positions: climbing, dropping, new entries
  4. Highlights strike zone keywords ready to push
  5. Flags content losing ground (act before it falls further)
  6. Saves new snapshot for next comparison

Snapshots stored at: ~/clawd/workspace/seo-agent/snapshots/SITE-YYYY-MM-DD.json

seo-compete.sh — Find Their Keywords, Take Them

bash
./seo-compete.sh --site example.com --competitor competitor.com [--limit 30]

What it does:

  1. Uses DataForSEO to find all keywords your competitor ranks for
  2. Cross-references against your rankings
  3. Returns the gap — keywords they're winning that you're not targeting
  4. Scored by their position + search volume (easiest wins first)

This feeds directly back into DISCOVER as a seed list.


Configuration

On first use, create ~/clawd/workspace/seo-agent/config.yaml:

yaml
site: "sc-domain:example.com"
competitors:
  - competitor1.com
  - competitor2.com
content_directory: "./content"
monitor_frequency: "weekly"
target_positions: [5, 20]   # the strike zone
min_search_volume: 100

Scripts read this config automatically. Override any setting with CLI flags.


Setup

1. DataForSEO Credentials

bash
export DATAFORSEO_LOGIN=your@email.com
export DATAFORSEO_PASSWORD=yourpassword

2. Google Search Console Auth

Same auth as gsc-report. If you've set that up, you're ready.

Options:

  • export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json
  • gcloud auth application-default login
  • export GOOGLE_ACCESS_TOKEN=ya29.xxx

3. Your Site URL

Find your exact property URL in Search Console. It's either:

  • https://example.com (URL prefix property)
  • sc-domain:example.com (domain property — broader, recommended)

4. (Optional) Brand Voice

Drop your brand voice profile at ~/clawd/workspace/brand/voice-profile.md. The agent reads this when generating content briefs and drafts. Run the brand-voice skill first if you haven't.


Running the Full Loop

Week 1 — Discovery

Run: seo-discover.sh --site your-site --limit 20
Output: Prioritized keyword hit list
Action: Pick top 3 opportunities → generate content with Claude

Week 2+ — Monitor + Optimize

Run: seo-monitor.sh --site your-site
Output: What moved, what's in strike zone
Action: Find climbers → write supporting content → add internal links

Ongoing — Compete

Run: seo-compete.sh --site your-site --competitor their-site
Output: Their keywords, your gaps
Action: Feed gaps back into discovery as seeds

How Claude Uses This

When you run the SEO Agent loop, Claude:

  1. Reads discovery output → selects the best opportunity based on your goals
  2. Generates a full content brief: target keyword, secondary keywords, outline, angle, word count, FAQ section, schema markup suggestions
  3. Drafts the article (or hands off outline for your writer)
  4. Reads monitor output → identifies what to optimize next
  5. Writes supporting articles with internal link strategy built in
  6. Updates workspace/brand/learnings.md with what's working

DataForSEO Endpoints Used

PurposeEndpoint
Keyword volumePOST /v3/keywords_data/google_ads/search_volume/live
Related keywordsPOST /v3/keywords_data/google_ads/keywords_for_keywords/live
Keyword expansionPOST /v3/dataforseo_labs/google/keyword_suggestions/live
SERP checkPOST /v3/serp/google/organic/live/advanced
Competitor keywordsPOST /v3/dataforseo_labs/google/competitors_domain/live
Keyword gapsPOST /v3/dataforseo_labs/google/domain_intersection/live

API base: https://api.dataforseo.com Auth: HTTP Basic — $DATAFORSEO_LOGIN:$DATAFORSEO_PASSWORD


Output Files

PathContents
workspace/seo-agent/snapshots/SITE-DATE.jsonWeekly ranking snapshots
workspace/seo-agent/config.yamlYour site config
workspace/brand/learnings.mdRanking insights (appended)
workspace/brand/keyword-plan.mdActive keyword targets

Why This Works

Most people do SEO once and forget it. They publish and pray.

The SEO Agent turns SEO into a system. Every week you know:

  • What moved
  • Why it moved
  • What to do next

Compound that for 6 months and you've got a content moat competitors can't buy their way out of.

The scripts are the engine. You (and Claude) are the brain. Together, that's the loop.

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 Agent AI skill do?

Self-improving SEO system. Finds keywords via DataForSEO, writes content that ranks, monitors GSC for what's climbing, then writes MORE content to push winners higher. A feedback loop, not a one-shot tool.

Why use Seo Agent on TypingMind?

Because you install it once and use it with any model. Seo Agent 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 Agent in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/TheMattBerman/seo-kit/tree/main/skills/seo-agent. 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 Agent?

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 Agent?

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

Is the Seo Agent AI skill free?

It is published on GitHub by TheMattBerman. Check the repository for licensing terms. 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 👇