Connect
Authorize a compatible MCP client with your member account and the narrow mcp:use scope, so an AI agent can retrieve the SEO data your account is allowed to inspect.
SEO data your AI agent can actually inspect
SEO MCP connects AI agents to real SEO evidence.
Use the Model Context Protocol to investigate Screpy crawls, links, images, tracked rankings, AI visibility, uptime, and Core Web Vitals. Ask a focused question, inspect the records, and choose the next action.
OAuth + PKCECrawls + rankings + AI visibilityPermission-awareStructured output
SEO MCP
connectedRequest
Find broken internal links in the latest crawl and prioritize the fixes.
Tools
list_projectslist_crawlslist_links
Illustrative result
A broken destination appears across a navigation template. The agent groups its source pages, checks the response, and suggests one shared fix with a follow-up crawl to verify it.
One connection, focused context
What does SEO MCP do?
SEO MCP gives compatible AI clients a standard, controlled way to retrieve SEO evidence from your projects. An agent can move from a broad question to exact project, crawl, page, link, image, ranking, AI answer, uptime, and Core Web Vitals records without losing the context behind the answer.
Ask
Describe an SEO outcome in plain language—such as finding 4xx pages, tracing broken links, or comparing releases—and let the client choose the structured tools.
Verify
Check the source, date, scope, and representative records behind each finding before turning it into a prioritized SEO action list.
Find the right SEO MCP path
Start with a connection, inspect a tool, or follow a complete investigation. Each section links to the evidence and next step needed for its task.
Get startedConnect a compatible AI client, authorize access, and verify the first read-only result.ToolsSee what each Screpy dataset can answer and which actions change account state.WorkflowsChoose a task and follow its evidence, interpretation, and validation steps.GuidesMake better prompting, content, automation, and MCP-versus-API decisions.
Built for technical SEO work
From a site-wide signal to the exact page behind it
Broad summaries are useful, but verifiable records make SEO decisions safer. SEO MCP connects crawl records, tracked rankings, and monitored AI answers to focused investigations. Each workflow keeps its dataset and limits visible.
Technical SEO audits
Inspect HTTP status codes, title tags, meta descriptions, heading structure, canonicals, structured data, crawl depth, response times, and indexability signals in one evidence-backed audit. Use filters to move from a site-wide pattern to the exact URLs that need attention.
Link intelligence
Find broken internal and external destinations, review anchor context, and trace each link back to its source pages. Prioritize navigation fixes, recover internal link equity, and separate template-wide issues from one-off content errors.
Image SEO analysis
Filter crawled images by file size, format, HTTP status, missing alt text, and source page. Build a focused image SEO queue for accessibility, performance, and indexability improvements without sorting through an unbounded export.
Core Web Vitals
Bring LCP, INP, CLS, and related page-experience signals into the same workflow as crawl and on-page data. Compare affected URLs with their templates and technical findings so performance work has clear SEO context.
Crawl comparisons
Compare crawls or projects to identify new errors, resolved issues, URL changes, and regression patterns. Use the differences to explain release impact and turn a large crawl history into a prioritized technical SEO backlog.
Permission-aware actions
Start with read-only project and crawl analysis, with access checked against the authenticated member on every tool call. State-changing operations remain explicit and require a separate confirmation before high-impact work can proceed.
Tracked keyword trends
Compare stored Rank Tracker analyses for the same keyword, location, and device. Inspect position and landing-page changes before proposing a content or technical explanation.
AI Visibility evidence
Review monitored prompts, brand mentions, answer text, and cited sources. Distinguish an observed mention from a cited page and keep model and date context attached.
SEO quick wins
Review bounded opportunities from available crawl, search, and connected traffic signals. Check the underlying records and time window before turning a suggestion into work.
Practical SEO MCP examples
Ask for the outcome, not the export.
Each workflow starts with a question an SEO practitioner would actually ask, then turns that question into a traceable sequence of structured tool calls. Project and dataset context stays attached while the agent finds evidence and explains the next action.
Explore all workflowsDiagnose a traffic-critical issue
“Find every 4xx page in my latest crawl, group the issues by impact, and show which URL patterns or templates should be fixed first.”
list_projects → list_crawls → list_pages
Recover internal link equity
“Show broken internal links in my latest crawl, include their source pages and anchor context, and prioritize fixes that affect the most important site sections.”
list_links → get_link_sources
Measure release impact
“Compare my two latest crawls, summarize resolved and new SEO issues, and explain the regressions that could affect organic visibility after the release.”
list_crawls → compare_crawls
Build an image optimization queue
“List large images without alt text, include every page where each image appears, and group the results into an actionable image optimization queue.”
list_images → get_image_sources
Explain a tracked ranking change
“Compare the latest stored analyses for a keyword on the same device and in the same location. Show dates, positions, and landing URLs.”
get_rank_tracker_overview → get_rank_tracker_keyword
Inspect AI answer visibility
“Show monitored prompts that mention my brand, then inspect representative answers and cited sources. Keep model and date context.”
get_ai_visibility_overview → list_ai_visibility_prompts → get_ai_visibility_answer
Access with guardrails
Your account. Your projects. Explicit control.
Access is scoped to the authenticated member and checked on every call. The server directs agents to discover valid project and crawl identifiers instead of guessing, while state-changing operations stay separate from normal read-only SEO analysis.
- Project access rechecked for every tool call
- Structured filters instead of unrestricted queries
- Explicit confirmation for high-impact actions
- Cursor pagination and bounded result sizes
Connection details
- Endpoint
- https://mcp.screpy.com
- Authorization
- OAuth authorization code + PKCE
- Scope
- mcp:use
- Default
- Read-only analysis
Common questions
SEO MCP FAQ
What is SEO MCP?
+
SEO MCP is a Model Context Protocol server that gives compatible AI agents permission-aware access to Screpy project, crawl, page, link, image, uptime, Core Web Vitals, Rank Tracker, and AI Visibility data.
Does SEO MCP replace an SEO crawler?
+
No. SEO MCP connects an AI agent to structured crawl and monitoring data. The crawler collects evidence; the agent helps you query, compare, explain, and act on that evidence.
Can an AI agent change my projects automatically?
+
Read-only analysis is the normal path. Changing projects, starting crawls, or adding tracked keywords and prompts requires an explicit user request. High-impact actions also require confirmation.
Which SEO tasks can I automate?
+
Common workflows include technical audits, broken-link and image reviews, crawl comparisons, quick-win discovery, ranking trends, AI visibility, uptime, and Core Web Vitals reviews.
How does authentication work?
+
The member SEO MCP server uses an OAuth authorization-code flow with PKCE and the mcp:use scope. Access stays tied to the authenticated member and is rechecked for every tool call.
Is SEO MCP useful for AI search optimization?
+
Yes. It can audit technical foundations and inspect monitored AI answers, brand mentions, and cited sources. These observations do not guarantee citations or establish what caused them.
Give your AI agent real SEO context.
Start with the connection guide, run your first evidence-backed SEO audit, then build repeatable workflows around crawl health, rankings, AI visibility, internal links, images, performance, and releases.