A side-by-side look at the Citation Intelligence and Arxiv MCP Server MCP servers: transport, auth, tools, quality signals, and the config to install either one in Claude, Cursor, Windsurf or VS Code.
Verdict: choose Citation Intelligence for search & data extraction work as a local process; choose Arxiv MCP Server for search & data extraction as a local process. You can also run both side by side in one mcpServers block.
SERPAPI_KEY, BING_WEBMASTER_API_KEY, GOOGLE_APPLICATION_CREDENTIALS.citations_check, domain_am_i_cited, signals_ai_overview.search_papers, download_paper, list_papers.Rows where the two servers differ are marked.
| Spec | ||
|---|---|---|
| Category | Search & Data Extraction | Search & Data Extraction |
| Quality signal | 57/100 · Good | 69/100 · Great |
| Install command | Confirmed | Confirmed |
| Transport | Local subprocess (stdio) | Local subprocess (stdio) |
| Auth | API key required | No auth required |
| Pricing | BYOK (pay the provider) | Free |
| Required env vars | SERPAPI_KEYBING_WEBMASTER_API_KEYGOOGLE_APPLICATION_CREDENTIALS | None listed |
| Verified owner | Community listing | Community listing |
| GitHub stars | 9 | 3,211 |
| Directory activity | 2 views 0 installs 0 upvotes | 3 views 0 installs 0 upvotes |
| Tools | 26 | 19 |
| Full listing | View Citation Intelligence | View Arxiv MCP Server |
citations_checkReturn URLs cited by an AI engine (Perplexity, Claude, ChatGPT, Gemini, or Bing) for a query. Use this when an agent or user wants to see what sources an AI search engine grounds answers on. Requires at least one engine API key; auto-picks the first available.domain_am_i_citedCheck whether a domain is cited by an AI engine across a cluster of queries. Returns per-query presence, rank, and a citation-rate summary. Use to measure visibility for a brand, product, or content site in AI search.signals_ai_overviewCheck whether Google shows an AI Overview for a query, and which URLs it cites. Uses SerpAPI (free tier: 100/month). Set SERPAPI_KEY.domain_cited_forList queries that the given domain has been cited for, served from the local cache. Build up a corpus by calling check_citations or am_i_cited first; cited_for queries it without spending API budget.citations_predictScore citation likelihood for a URL from public signals (Wikipedia link presence, schema.org markup, /llms.txt, GitHub and Reddit references, canonical hygiene, HTTPS). No LLM fired - all heuristic. Returns 0-100 score, grade, signal breakdown, and ranked fixes.panel_trackSave, load, or list named query panels. A panel is a persisted set of queries you want to monitor over time (e.g. editorial-watchlist). Use action=save with queries[] to create, action=load to read, action=list to enumerate. Panels live under <config>/panels/<name>.json.panel_runRun a saved panel through am_i_cited and append a timestamped snapshot. Side effects: makes external API calls to the configured AI engine (costs API quota); writes one snapshot file to disk at <config>/snapshots/<panel>/<iso>.json. Requires at least one engine API key (same as am_i_cited). Returns per-query citation presence and a citation_rate summary for the run. Use panel_track to create a panel first; use citations_trend to read the accumulated trend after multiple runs.report_visibilityTurnkey AI visibility report for a domain across a query set. Composes check_citations over every query (or a saved panel) and returns the metrics AI-visibility trackers sell as a dashboard, in one call: mention frequency (citation_rate), share_of_voice vs competitors, average rank when cited, and brand sentiment from the answer text. Side effects: one check_citations call per query (costs API quota for uncached queries; cached queries are free). Returns structured summary + top_domains + per_query, plus a rendered Markdown report (include_markdown=true) suitable for a public page. Provide queries[] or a panel name. Same engine selection as check_citations.citations_trendReport citation rate over time for a panel from stored snapshots. Read-only; cache-only — makes no API calls to any AI engine and costs no API quota. Reads snapshot files from <config>/snapshots/<panel>/. Returns: snapshots[] (one entry per panel_run invocation, each with timestamp and citation_rate), plus per-query deltas (gained/lost/unchanged) comparing first vs last snapshot. Returns an empty series when no snapshots exist yet. No auth required. No rate limits. Use panel_run to accumulate snapshots first; use since to restrict the time window.competitors_compareRun predict_citation on 2-10 URLs and return a side-by-side signal table plus a list of signals where the URLs diverge. Use to compare your URL to top-cited competitors for the same query.signals_wikipediaList Wikipedia articles that reference the given domain. Read-only. One HTTPS GET to the Wikipedia API (en.wikipedia.org/w/api.php?action=query&list=exturlusage). No auth required; no API keys; no rate limits beyond Wikipedia's public API fair-use policy (~1 request/second). Returns article titles and URLs. Wikipedia backlinks are the highest-lift signal for LLM training corpora — a domain cited from Wikipedia is far more likely to appear in AI training data and citation pools. Use lang to query non-English Wikipedias.audit_sitemapFetch a sitemap.xml (or sitemap index) and run predict_citation on every URL. Returns results sorted worst-score-first. Surfaces systemic issues across a whole site in one pass. Zero engine keys needed.search_papersSearch arXiv by query with optional categories, date range, sort, and pagination.
Query: prefer quoted phrases; ti:/au:/abs:/cat:; AND/OR/ANDNOT. Unprefixed terms match title+abstract (not authors). Use categories (cs.AI, cs.LG, cs.CL, cs.CV, cs.MA, cs.RO, stat.ML, quant-ph). Catalog/examples: README 'search_papers query guide'.
Dates YYYY-MM-DD (date_from/date_to). sort_by relevance|date. max_results default 5 (cap 50). abstract_mode none|snippet|full (default snippet). start default 0; response: total_results, returned, has_more, next_start, abstract_mode. Pass next_start with same abstract_mode. Use get_abstract after compact search — not after abstract_mode=full.
arXiv ~3s between requests (server-side). Transient 429/503 are retried with backoff; persistent rate limits return status=rate_limited.download_paperDownload a paper from arXiv and return its text content. Tries the HTML version first for clean extraction; falls back to PDF conversion if HTML is unavailable. Stores the paper locally. Returned text is bounded to roughly 12,000 characters by default so one call cannot return an unbounded paper body. When is_truncated is true, call again with start=next_start (see next_retrieval) to continue, or pass return_full_text=true for the entire remaining paper. Set force=true to re-fetch and overwrite a cached paper (required to replace a newer stored arXiv version with an older one).list_papersList all papers that have been downloaded and stored locally via download_paper. Returns id, title, authors, published, and arxiv_version/versioned_id from local metadata — no live re-fetch. Set compact=true to return arXiv IDs only. Returns an empty list if no papers have been downloaded yet. Workflow: search_papers -> download_paper -> list_papers -> read_paper.read_paperRead the text content of a paper that was previously downloaded via download_paper. Returns the paper in markdown format, bounded to roughly 12,000 characters by default so one call cannot return an unbounded paper body. When is_truncated is true, call again with start=next_start (see next_retrieval) to continue, or pass return_full_text=true for the entire remaining paper. Will fail with a clear error if the paper has not been downloaded yet — call download_paper first. Workflow: search_papers -> download_paper -> read_paper.get_abstractFetch abstract and metadata by arXiv ID without downloading the paper. Use before download_paper to assess relevance. Returns title, authors, abstract, categories, published date, and PDF URL. After compact search, use for one full abstract; skip if search used abstract_mode=full.Paste either block (or both) into your client config, e.g. claude_desktop_config.json or ~/.cursor/mcp.json. We only show a config when the install command is confirmed.
{
"mcpServers": {
"automatelab-tech-citation-intelligence": {
"command": "npx",
"args": [
"-y",
"@automatelab/citation-intelligence"
],
"env": {
"SERPAPI_KEY": "YOUR_SERPAPI_KEY_HERE",
"BING_WEBMASTER_API_KEY": "YOUR_BING_WEBMASTER_API_KEY_HERE",
"GOOGLE_APPLICATION_CREDENTIALS": "YOUR_GOOGLE_APPLICATION_CREDENTIALS_HERE"
}
}
}
}
{
"mcpServers": {
"blazickjp-arxiv-mcp-server": {
"command": "uvx",
"args": [
"arxiv-mcp-server"
]
}
}
}
semantic_searchSemantic similarity search over papers you have already downloaded locally via download_paper. Supports free-text queries (e.g. 'attention mechanisms for long sequences') or finding papers similar to a given paper_id. IMPORTANT: only searches your local downloaded collection — will return empty results if no papers have been downloaded yet. Use search_papers to find papers on arXiv, then download_paper to add them to the local index before using this tool. Requires pro dependencies: uvx --from 'arxiv-mcp-server[pro]' arxiv-mcp-server (or uv tool install 'arxiv-mcp-server[pro]').reindexRebuild the local semantic index for downloaded papers.citation_graphReturn papers citing an arXiv paper and papers that it references using Semantic Scholar's citation graph. Results are bounded (default 50) to stay within the unauthenticated quota. Under load, export SEMANTIC_SCHOLAR_API_KEY for a higher limit; without a key, persistent rate limits return status=rate_limited instead of failing hard.export_citationsExport BibTeX citations for one or more arXiv papers using authoritative arXiv metadata (title, authors, year, primary category), never model-generated fields. Version suffixes (e.g. '2401.12345v2') are preserved and citation keys are deterministic. Returns the rendered BibTeX plus per-paper status/error. BibTeX only; RIS/CSL-JSON are not yet supported.watch_topicSave or update a persistent research topic watch. When checked via check_alerts, returns only papers published since the last check — acting as a standing alert for new work on a topic. New watches seed last_checked to creation time so the first check does not dump historical matches. The topic string uses the same query syntax as search_papers (quoted phrases, field specifiers, boolean operators). Examples: '"diffusion models" AND ti:"video generation"', 'au:"LeCun" AND cs.LG'. Calling watch_topic with the same topic string updates the existing watch rather than creating a duplicate. On update, omit categories to preserve existing filters; pass categories: [] to clear them. Pair with check_alerts to poll for new papers.check_alertsCheck all saved topic watches for newly published papers since the last check. Omitting the topic parameter runs ALL saved watches and returns new papers for each. Passing a topic string checks only that specific watch. Advances each watch's drain cursor after running: when a page is truncated by max_results, has_more=true and check_start advances so later calls return the next papers in the same window (Atom date bounds alone are day-granular and would otherwise re-hit the boundary); last_checked tracks the newest returned paper. When the page is not full, last_checked becomes now and the drain cursor resets. Use watch_topic to register topics before calling this. Returns a clear not-found error if a topic is provided but no matching watch exists. Returns a summary with new paper counts, has_more, and full paper metadata per topic.list_watchesList all saved topic watches without checking for new papers. Returns each watch's topic, categories, last_checked timestamp, and other stored fields. Does not update last_checked — use this to inspect what is saved. Use unwatch_topic to remove a watch, or check_alerts to poll for new papers.