data-aggregator-mcpSAFE
Unified research-data acquisition MCP — search & fetch datasets across Zenodo, DataCite, NCBI omics (GEO/SRA/BioProject), and literature (PubMed/OpenAIRE) behind one normalized model.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Search 18 research-data sources at once (data archives, omics repositories and papers) and get one list back. Organism, disease and tissue names are expanded with their synonyms in the 12 sources that accept them, records that share a DOI are collapsed to one, and downloads are checked against the source's checksum where it publishes one.
[](https://pypi.org/project/data-aggregator-mcp/) [](https://pypi.org/project/data-aggregator-mcp/) [](https://github.com/musharna/data-aggregator-mcp/blob/main/LICENSE) [](https://github.com/musharna/data-aggregator-mcp/actions/workflows/ci.yml) [](https://glama.ai/mcp/servers/musharna/data-aggregator-mcp) [](https://doi.org/10.5281/zenodo.21636332)
[](https://insiders.vscode.dev/redirect/mcp/install?name=data-aggregator&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22data-aggregator-mcp%22%5D%7D) [](https://cursor.com/install-mcp?name=data-aggregator&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJkYXRhLWFnZ3JlZ2F0b3ItbWNwIl19)
mcp-name: io.github.musharna/data-aggregator-mcp
46cf60095b50OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add data-aggregator-mcp --env NCBI_API_KEY=${NCBI_API_KEY} --env LLM_API_KEY=${LLM_API_KEY} --env EMBEDDING_API_KEY=${EMBEDDING_API_KEY} --env DATA_GOV_API_KEY=${DATA_GOV_API_KEY} -- uvx data-aggregator-mcp==0.69.1Trust audit
SAFEgrade C · trust 73/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (5 observation(s))
- Shell
- declared (5 observation(s))
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (25)
sample.parquet
.pre-commit-config.yaml
.zenodo.json
check bandit-medium-plus sec1.py $'import hashlib\nh = hashlib.md5(b"x").hexdigest()' fail
check bandit-medium-plus sec2.py $'import hashlib\nh = hashlib.md5(b"x", usedforsecurity=False).hexdigest()' pass
check bandit-medium-plus sec3.py $'import hashlib\nh = hashlib.md5(b"x").hexdigest() # nosec B324 - content fingerprint an external tool also computes' pass
digest = hashlib.sha1(data, usedforsecurity=False) # pragma: no mutate
assert f.checksum == f"md5:{hashlib.md5(body).hexdigest()}" # nosec B324 - md5 is the upstream's checksum algorithm, compared not trustedreturn _res([FileEntry(name="secrets.parquet", url="file:///home/user/.ssh/id_rsa")])
[(tarfile.SYMTYPE, "/etc/passwd"), (tarfile.LNKTYPE, "plain.txt")],
PARQUET_URL, "sample.parquet", "SELECT * FROM read_csv_auto('/etc/passwd')"PARQUET_URL, "sample.parquet", "SELECT * FROM read_csv_auto('/etc/passwd')"@pytest.mark.parametrize("bad", ["../escaped.txt", "a/../../escaped.txt", "a/..", "."])await biostudies.search(c, "x", collection="../../etc/passwd")
@pytest.mark.parametrize("bad", ["biostudies:", "biostudies:../../secret", "biostudies:a/b"])"a dot-dot segment": _with("fastq_ftp", f"ftp.sra.ebi.ac.uk/vol1/../../x.gz;{_TWO}"),assert await ena.filereport(c, "../../evil?injected=1&format=xml#x") == []
cls = zipfile.ZipExtFile if kind == "zip" else tarfile.ExFileObject
id_rsa ... httpx.
os.environ.items() ... urllib.request
`http://169.254.169.254/` was fetched, and with `op='head'` the body came back to the
("http://169.254.169.254/latest/meta-data/", "link-local: cloud metadata"),URL that 302s to `http://127.0.0.1:9200/` reached it with the check satisfied. Measured
whose file is _named_ `data.csv` while pointing at `http://127.0.0.1:9200/` or
`http://169.254.169.254/` was fetched, and with `op='head'` the body came back to the
Gates applied: no_behavioural_pass.
46cf60095b50full audit observations/trust-audit/mcp-server/musharna__data-aggregator-mcp.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 46cf60095b50 | SAFE | C | 73 | first audit |
Questions
What is the data-aggregator-mcp MCP server?
Unified research-data acquisition MCP — search & fetch datasets across Zenodo, DataCite, NCBI omics (GEO/SRA/BioProject), and literature (PubMed/OpenAIRE) behind one normalized model.
Is data-aggregator-mcp safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it C (73/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does data-aggregator-mcp need?
It reads DATA_GOV_API_KEY, EMBEDDING_API_KEY, LLM_API_KEY and NCBI_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does data-aggregator-mcp run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on PyPI as data-aggregator-mcp.
How current is this page?
The grade is for one exact copy of the source (46cf60095b50), read on 2026-10-11. The repository is watched and re-audited when it changes.