
SEO research and site audits for your coding agent. 10 agent skills + a local CLI for keyword
SEO research and site audits for your coding agent. 10 agent skills + a local CLI for keyword research, competitor analysis, backlinks, local SEO and rank tracking. No server, pay per lookup.
·
SEO research and site audits for your coding agent.
AgenticSEO gives an agent such as Pi or Claude Code a command-line toolkit and ten ready-made SEO workflows (Agent Skills). Ask it to audit your site, find keywords worth targeting, study a competitor or check your Google Maps visibility. It gathers the data, explains what matters, and saves a report in your project.
AgenticSEO brings the SEO features of OpenSEO to local commands and agent skills. It is an independent project, not an official OpenSEO release.
You need Node.js 24.14 or newer and a DataForSEO account (pay per lookup, no subscription).
# 1. Install the CLI. Run it again to update.
npm install -g https://github.com/kid7st/agenticseo/releases/latest/download/agenticseo.tgz
# 2. Add the skills to your agent. The installer asks which agents:
# Pi, Claude Code, Codex, Cursor and many more. Run it again after updating.
npx skills add "$(npm root -g)/agenticseo" -g
# 3. Give it your DataForSEO login and API password (base64-encoded).
# Put this line in your shell profile to keep it.
export DATAFORSEO_API_KEY="$(printf '[email protected]:your-api-password' | base64)"For ongoing work, start with "Set up SEO for this site": the seo-project-setup skill records your business, goals and competitors so later requests use them, and connects Google Search Console if you want. Not sure what to ask? Ask for the seo-coach.
So far AgenticSEO has been tested with Pi. If you use another agent, please tell us how it went.
We audited fastagent.sh, a TypeScript agent framework, starting from an empty folder:
mkdir fastagent.sh && cd fastagent.sh
agenticseo init --domain fastagent.sh --location US # the site and its market
pi # or your own agentAudit this site and tell me what to fix first. Work without asking me questions.
In about six minutes the agent:
.agenticseo/reports/.What it found: fastagent.sh ranks #13 (page 2) for its own name, "fastagent" (260 US searches a month), behind three unrelated projects with the same name, and it has only 2 referring domains. What to fix first: say "TypeScript" in the homepage title and description so searchers can tell it apart, earn a few developer links, and connect Search Console to see real clicks.
Read the full report, with every query, position and cost behind it.
| Ask your agent to… | Skill | Cost in our tests |
|---|---|---|
| Audit my site and tell me what to fix first | seo-audit | $0.17 (the crawl is free) |
| Find keywords worth targeting for these topics | keyword-research | under $0.30 |
| Group my keywords and map them to pages | keyword-clustering | $0.05 |
| Analyze this competitor | competitor-analysis | $0.20 |
| Show who wins my search market and where the openings are | competitive-landscape | $0.65 |
| Check how I show up on Google Maps | local-seo | $0.13 |
| Find sites that might link to my guide | link-prospecting | $0.19 |
Each workflow ends with a report: a Markdown file your agent can read later and an HTML page you can open, share or print. Costs are what DataForSEO charged in one test run; you can set a budget in your request ("keep it under $0.50"). See skills for details.
The CLI also tracks keyword rankings on demand or on a schedule, and shows your own Search Console and Analytics data for free.
you ──ask──▶ your agent ──reads──▶ skill (the workflow)
│
└──runs──▶ agenticseo CLI ──▶ DataForSEO · Google Search Console · Analytics · local crawler
│
└──▶ .agenticseo/ results, raw evidence, reportsSkills tell the agent what to ask you, which commands to run and how to read the results. The CLI does the data work and prints JSON. For example, agenticseo keywords "seo audit" "seo audit tool" returns:
{
"provider": "DataForSEO",
"market": { "locationCode": 2840, "languageCode": "en" },
"rows": [
{ "keyword": "seo audit", "searchVolume": 4400, "cpc": 16.06, "keywordDifficulty": 77, "intent": "commercial" },
{ "keyword": "seo audit tool", "searchVolume": 2400, "cpc": 23.77, "keywordDifficulty": 77, "intent": "commercial" }
],
"costUsd": 0.01224,
"evidence": ".agenticseo/evidence/2026-09-25T03-30-58.479Z-….json"
}A metric the provider does not have is null, never 0. You can run every command yourself; the command reference covers them all.
| Service | Needed for | Cost |
|---|---|---|
| DataForSEO | Keywords, SERPs, competitors, backlinks, local SEO, AI visibility, rank tracking, Lighthouse | Pay per lookup, usually $0.001–$0.08; an AI brand lookup is about $0.60–$0.85 |
| Google Search Console and Analytics | Your site's own clicks, queries and visitor behavior | Free; needs a one-time Google setup |
| Ahrefs (optional) | Ahrefs Domain Rating | Free API key |
Keys live in your environment and never go into project files. Repeated lookups come from a local cache and cost nothing.
Everything for a site stays in a .agenticseo/ folder inside it:
| Path | What it holds | In Git? |
|---|---|---|
project.json | Domain, market, and which Search Console and Analytics property to use | Your choice |
context.json | What your agent knows about the business: goals, positioning, competitors, key pages and a research log | Your choice |
reports/ | Finished reports (Markdown and HTML) | Your choice |
exports/ | CSV and JSON exports | Your choice |
evidence/ | The raw provider response behind every paid result | Ignored |
data/ | A SQLite database of saved keywords, audits and ranking history | Ignored |
cache/ | Cached lookups | Ignored |
Google access tokens are stored per user in ~/.config/agenticseo/, never in a project. Behind a proxy, set HTTPS_PROXY as you would for curl.
MIT. AgenticSEO includes source adapted from OpenSEO (MIT, Copyright (c) 2026 Ben Senescu). Each adapted file names its origin, and OpenSEO's notice is kept in LICENSES/OpenSEO.txt. AgenticSEO is not affiliated with OpenSEO's maintainers.
Paste this prompt into your agent. It reads this page and does the setup for you.
Read https://aiagentslisting.com/mcp/agenticseo to learn what the "AgenticSEO" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.Agents can also browse this directory over MCP at https://aiagentslisting.com/api/mcp
npx skills add "$(npm root -g)/agenticseo" -gnpm install -g https://github.com/kid7st/agenticseo/releases/latest/download/agenticseo.tgzThis server runs locally, so we can't read its tool list over the web yet.
We could not run this server: it ships no npx package for our sandbox and no public endpoint. The four checks that need a live server are left out, and the grade tops out at B until one exists.
Maintain this server? Claiming proves you own the listing and earns the last 4 points.
Claim this listingPaste this line near the top of your repository's README. The badge always renders the current score.
[](https://aiagentslisting.com/mcp/agenticseo)Nothing comparable is listed yet.
v0.2.0Release notesv0.1.0Release notesConnect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "agenticseo" with a rating from 1 to 5 and a short note on what worked and what did not.No reviews yet
Be the first to share your experience with this tool.
One email a week. New agents, MCP servers and skills, and what is actually getting traction.