Skip to main content

aletheia-mcp

An MCP server that gives AI agents structured access to scientific literature — 250M+ papers indexed via OpenAlex, covering arXiv, PubMed, Crossref, and more — so they can do research, not just search.

The gap this fills

Tools like Elicit, Consensus, and SciSpace are great but they're standalone chat apps. You ask one question, you get one answer. You can't chain operations.

With aletheia-mcp running in Claude or any MCP-compatible client, you can ask things like:

"I'm reviewing literature on mitochondrial dysfunction in Parkinson's. Find the 20 most-cited papers from the last 5 years, pull the full text of the three most influential, identify the exact mechanisms they propose, and flag any that contradict each other."

Under the hood, Claude chains multiple tool calls — search, fetch metadata, walk citations, download and parse full PDFs — and synthesizes. That's the difference between search and research.

Tools in v0.3.0

  • search_papers(query, limit) — unified search across the OpenAlex corpus
  • get_paper_details(paper_id) — full metadata for any paper by OpenAlex ID, DOI, PubMed ID, or URL
  • get_citation_graph(paper_id, direction, limit) — walk citations forward (direction="citations") or references backward (direction="references")
  • get_paper_full_text(paper_id) — new in v0.3 — fetches and extracts the full text of any open-access paper. Handles arXiv, PubMed Central, and direct publisher PDFs. Falls back gracefully when no OA version is available.

All paper-returning tools use a normalized shape: {id, title, abstract, authors, year, citation_count, venue, doi, url, pdf_url, source}. The full-text tool adds text and char_count.

Why OpenAlex (not Semantic Scholar)

Earlier iterations used Semantic Scholar. The free API key there is gated and institution-preferred — independent developers face slow or rejected applications, which meant future users of this tool would hit the same wall.

OpenAlex is the open, CC0-licensed successor to Microsoft Academic Graph, with 250M+ works covering the same sources. As of February 2026, API keys are required but are free and self-serve — 30 seconds to sign up, no approval process.

Install

Not yet on PyPI. To run locally:

git clone https://github.com/TechFusionData/aletheia-mcp
cd aletheia-mcp
uv sync
cp .env.example .env
# edit .env and paste your OpenAlex API key

Then register with Claude Code (or your MCP client of choice):

claude mcp add --scope user aletheia -- uv run --directory $(pwd) aletheia-mcp

Restart your MCP client and the tools become available.

Dependencies worth noting

  • PyMuPDF (AGPL-3.0) — used for full-text PDF extraction because it handles two-column scientific layouts far better than alternatives. If AGPL is a concern for your use case, you can swap in pypdfium2 (Apache-2.0) in sources/openalex.py:_extract_pdf_text at the cost of somewhat messier text output.

Why "aletheia"

ἀλήθεια — the Greek notion of truth as "unconcealment," the state of not being hidden. Research is the practice of bringing what's there into view.

License

MIT. See LICENSE.

Metadata

Release files for aletheia-mcp 0.3.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for aletheia-mcp 0.3.0
File Size Uploaded Provenance Actions
aletheia_mcp-0.3.0.tar.gz 13.0 kB Without provenance Details for aletheia_mcp-0.3.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for aletheia-mcp 0.3.0
File Interpreter ABI Platform Actions
aletheia_mcp-0.3.0-py3-none-any.whl Python 3 none any Details for aletheia_mcp-0.3.0-py3-none-any.whl

Total release size: 30.0 kB

Release files / aletheia_mcp-0.3.0.tar.gz

Download URL aletheia_mcp-0.3.0.tar.gz
Size 13.0 kB
Tags Source
SHA-256 checksum
How to use checksums
17b9dcf07a7a98022911741a6d4c3e5e4ce35e624cc38f2cfbbd41d7f428303f
BLAKE2b-256 checksum
How to use checksums
f991b90baf2d7d1312a61c4d616fc1226ca3f729e492a4d7e230b01da5226324
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.11.7 {"installer":{"name":"uv","version":"0.11.7","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Release files / aletheia_mcp-0.3.0-py3-none-any.whl

Download URL aletheia_mcp-0.3.0-py3-none-any.whl
Size 17.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
914fc9a75e12c95bf99d517c9fc7ff539e3ff3830f0ed35bdeb3ae67321791aa
BLAKE2b-256 checksum
How to use checksums
ca48676eb1e51ca7556ab51f4191021ad3533f912b71212a6be52dff1fae7b0a
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.11.7 {"installer":{"name":"uv","version":"0.11.7","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

Security for aletheia-mcp 0.3.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.3.0 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page