Science AI Journal

Developers · MCP Server

Call Science AI Journal tools from Claude Desktop, Cursor & Cline

Install one npm package and the six Science AI Journal tools appear inside your MCP-enabled assistant. No more switching tabs to upload PDFs — Pre-Check a paper, find target journals, run duplicate checks, surface research gaps, or drive the article writer right where you're working.

Generate API key →View on npm

Quick start (≈ 2 min)

  1. 1

    Generate an API key

    In your Settings → API Keys page, click New API key, name it (e.g. Claude Desktop on laptop), copy the saij_… token. You'll only see it once.

  2. 2

    Add the MCP server to your client config

    Drop this block into your Claude Desktop config file:

    {
  "mcpServers": {
    "science-ai-journal": {
      "command": "npx",
      "args": ["-y", "science-ai-mcp-server"],
      "env": {
        "SCIENCE_AI_API_KEY": "saij_…paste your key here…"
      }
    }
  }
}

    Config file location:
    macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    Windows: %APPDATA%\Claude\claude_desktop_config.json
    Linux: ~/.config/Claude/claude_desktop_config.json

  3. 3

    Restart your client and try a tool

    Quit and reopen Claude Desktop (or Cursor / Cline). When it comes back, the science-ai-journal server appears in the 🔌 tool list. Try this prompt:

    Use science-ai pre_check_paper on this paper:
Title: Robust Gradient Estimation for Sparse Cross-Modal Transformers
Abstract: We propose a new gradient-estimation scheme for sparse cross-modal transformers…

Tool catalogue

pre_check_paperFree — zero LLM cost.

Tier 1-5 acceptance probability from title + abstract.

When to use: Fast sanity-check before submission. Returns a predicted tier, field detection, and a confidence band — all from our local 33,000-paper FTS5 corpus, zero LLM call.

Example: “Use science-ai pre_check_paper on this abstract: "Title… Abstract…"”

check_duplicate_publicationFree — public-API lookups only.

Cross-reference for prior publication + salami slicing.

When to use: Pre-submission duplicate check across CrossRef, arXiv, medRxiv, bioRxiv, Unpaywall, and our 900k-paper institutional library. ~30 seconds.

Example: “Check my paper for prior publication before I submit — title and abstract are…”

recommend_journalsFree — zero LLM cost.

Ranked target-journal shortlist from a 1,214-venue index.

When to use: Get 5-10 ranked journal recommendations with letter grades, OA status, citation rate, predatory flags, and example similar papers. Local BM25 + topic-RAG only.

Example: “Recommend 10 target journals for this paper and exclude any predatory venues.”

find_research_gapsUses your Science AI Journal credits.

Surfaced gaps + most-cited + most-recent papers in your field.

When to use: Discovery for early-stage research: returns synthesised gaps drawn from cross-paper analysis plus the most-cited and most-recent papers around your query.

Example: “Find current research gaps in graph neural networks for drug discovery.”

hakem_review_paperUses your Science AI Journal credits.

Single-agent peer-review pass on a manuscript prompt.

When to use: Run one of the HAKEM specialist agents (methodology, language, figures, plagiarism, literature) on a prepared prompt. Web UI is still the easiest way to run the full 5-agent + synthesis flow.

Example: “Run a HAKEM methodology review on this prepared agent prompt: …”

start_writer_pipeline / get_writer_pipeline_statusUses your Science AI Journal credits.

Drive the Article Writer pipeline from outside the wizard.

When to use: Already have a writer session? Kick off Step 7.5 (code-search → adapt → verify → simulate) or any other section from your assistant, then poll status until done.

Example: “Start the writer pipeline for session "cl…abc" on step_7_5, then poll until status is "done".”

Other MCP clients

The same npm package works with every MCP-host that speaks stdio — only the config-file format differs.

Cursor

~/.cursor/mcp.json
{
  "mcpServers": {
    "science-ai-journal": {
      "command": "npx",
      "args": ["-y", "science-ai-mcp-server"],
      "env": { "SCIENCE_AI_API_KEY": "saij_…" }
    }
  }
}

Cline (VS Code)

cline_mcp_settings.json (via the Cline extension panel)
{
  "mcpServers": {
    "science-ai-journal": {
      "command": "npx",
      "args": ["-y", "science-ai-mcp-server"],
      "env": { "SCIENCE_AI_API_KEY": "saij_…" }
    }
  }
}

Continue / generic

Anywhere that supports an "mcpServers" object
{
  "mcpServers": {
    "science-ai-journal": {
      "command": "npx",
      "args": ["-y", "science-ai-mcp-server"],
      "env": { "SCIENCE_AI_API_KEY": "saij_…" }
    }
  }
}

Troubleshooting

Tools don't appear after editing config: fully quit the host app (not just close the window) and reopen. On macOS use ⌘Q.

401 unauthorized when calling a tool: the API key was wrong, revoked, or expired. Generate a fresh one and update the config.

429 too many requests / rate-limit on /api/research-gaps: you hit the 10-per-hour IP quota on research gaps. Wait an hour and retry, or move to a different IP.

npx complains it can't resolve the package: ensure Node ≥ 20 and that your npm registry isn't set to a private mirror that hasn't mirrored the package yet. Override with npm config set registry https://registry.npmjs.org/.

Frequently asked questions

The npm package and the local tools (pre_check_paper, recommend_journals, check_duplicate_publication) are free with zero per-call LLM cost. The agent-backed tools (hakem_review_paper, find_research_gaps, start_writer_pipeline) consume credits from your Science AI Journal account at the same rate as the web UI.

Ready to wire it up?

Generate your first key, paste the config, restart your client — two minutes from here to using Pre-Check inside Claude Desktop.

Generate API key →

Command palette

Jump anywhere, run any action.