Repolog MCP server

• Addresshttps://repolog.io/api/mcpnot in the official registry: answered when asked

No entry in the official MCP registry names this server. It answered on repolog.io when the Atlas asked at the address servers are usually mounted on, so it is the vendor's own. Clients and catalogues that list servers from the registry do not show it.

• Stateanswerschecked 2026-10-28
• Tools16
• Server calls itselfrepolog
• Server version1.0.0
• Protocol version2025-06-18

On 2026-10-28 the server answered initialize and tools/list without a sign-in. First answer in 46 ms.

• Its tools do every use case we checked: audit a homepage for SEO and speed issues; check accessibility failures and security headers on a site; map target keywords to pages and track them.
• Checked on 6 of the 7 days it takes to judge reliability.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

• INITE MCP scoreafter a weeka score needs 7 daily checks: 6 so far
• What an agent can do with it3 of 3 use cases
• Answeredmeasuring6 of the 7 daily checks it takes to print a share
• Median first answer2037 ms
• Space in an agent's context≈ 4,242 tokensthe definitions of all its tools (16), which an agent carries on every turn
• Tool definitionsB · 79 of 100
• Typed results0 of 16 toolsdeclared outputSchema: results come typed, read once

• Parameters without a description: in 4 of 16
• Not saying whether it only reads or can destroy: 16 of 16

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

• list_websites

  List all websites in the authenticated workspace, newest first.

• list_pages

  List the pages registered under a website. Each row's `id` is the pathId/targetPathId every other tool takes: pass it to run_scan to audit that page instead of the domain root, or to track_keyword/set_keyword_target_page to say which page…

• add_pages

  Register pages on a verified website so they can be scanned and targeted by keywords. Takes paths or absolute URLs on the website's own host — the site's sitemap is the usual source. Partial by design: `added` holds the registered pages…

• list_scans

  List scans for a website, newest first.

• get_scan

  Fetch a single scan by id. Defaults to the summary view: scores, per-category pass/warn/fail counts, the ranked top issues and — when the page has a keyword mapped to it — how it scored against that keyword. Enough to decide what to fix…

• run_scan

  Start a new scan for a website. An unverified website can scan its domain root (homepage) only; registered pages need domain verification. Returns immediately with a pending scan; poll get_scan or call wait_for_scan for the report.

• wait_for_scan

  Wait (up to ~45s) for a scan to finish, then return it in the same views as get_scan (summary by default). If it is still running when the wait elapses, returns the current pending state — call again to keep waiting. A scan of a slow page…

• list_keywords

  List the keywords tracked for a website, with search volume, difficulty, CPC and Search Console metrics where known. A null searchVolume with a set metricsUpdatedAt means a run checked and the provider has no data for that term — that is…

• get_search_console_queries

  Read Google Search Console for a website: the queries it already gets impressions for but isn't tracking (`opportunities`, best first, last 28 days, each with the `topPage` already earning them — the page to rewrite or to map the keyword…

• track_keyword

  Track a keyword for a website. Metrics stay empty until a research run prices it. Pass targetPathId to map it to a page — or targetRoot: true for the homepage, which has no page id of its own — and the next scan of that page grades it…

• untrack_keyword

  Stop tracking a keyword. Its metrics and page mapping are deleted with it. `keywordId` also takes an array to do several in one call; the other arguments then apply to every one of them.

• set_keyword_target_page

  Point a tracked keyword at the page it should rank for: a registered page by targetPathId, or the homepage with targetRoot: true (the root is always scannable, so it has no page id). Omit both to clear the mapping. The next scan of that…

• run_keyword_research

  Spend one of the workspace's monthly keyword research runs (pooled across every website) and write the results to the website's keyword list — at most 50 new keywords per run, shared across the seeds and ranked by how many words they…

• get_backlinks

  Backlink profile for a website: the latest snapshot (domain rank 0-1000, total backlinks, referring domains, broken backlinks, 0-100 spam score) plus up to two years of weekly history for trends. Snapshots are taken automatically about…

• get_fix_prompt

  Get a ready-to-use prompt describing what to fix in the code, built from a completed scan's report.

• get_usage

  The workspace's plan and how much of each quota is used and left: scans and keyword research runs (monthly, pooled across all websites, reset at `period.resetsAt`, UTC), websites, members, and tracked keywords (capped per website — see…