Database Lookup
Turns vague questions into reproducible API queries across 78 documented public databases, complete with provenance.
What it does
Instead of answering from general knowledge, this skill makes Claude retrieve facts from real public database APIs. It ships reference files for 78 sources across physics/astronomy (NASA, SIMBAD, SDSS), earth & environment (USGS, NOAA, EPA), chemistry & drugs (PubChem, ChEMBL, OpenFDA, BindingDB), materials (Materials Project, COD), biology & genomics (UniProt, Ensembl, PDB, AlphaFold, GEO, gnomAD), disease & clinical (Open Targets, ClinicalTrials.gov, ClinVar, GDC), patents & regulatory (USPTO, SEC EDGAR), and economics & demographics (FRED, BLS, World Bank, Census, WHO).
The emphasis is reproducibility: define a retrieval contract, pick the authoritative source, separate server-side from local filters, paginate exhaustively, reconcile counts, and report endpoints, parameters, access date and identifier conversions. It also codifies safety rules — treat API payloads as untrusted third-party data, never interpolate raw text into shell commands, never leak API keys.
Who it's for
- Researchers and grad students who need citable numbers with a traceable source
- Bioinformatics practitioners juggling gene, protein, variant and compound identifiers
- Analysts pulling repeatable macroeconomic or demographic time series
- Developers building hallucination-resistant fact-retrieval pipelines
Examples
- "Get UniProt function annotations and the AlphaFold model for TP53" → resolves the gene symbol to an NCBI Gene ID / UniProt accession, queries both APIs, and appends a provenance block.
- "Aspirin: PubChem CID, ChEMBL bioactivity, and FDA adverse-event summary" → follows name → CID → ChEMBL ID conversion and issues a bounded set of parallel calls.
- "All recruiting clinical trials for a rare disease" → paginates ClinicalTrials.gov by cursor, reconciles the reported total against retrieved records, and flags any incompleteness explicitly.
· · · Install guide · · ·
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on Claude Skill Mart. Copy the skills/database-lookup folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/database-lookup/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/K-Dense-AI/scientific-agent-skills.git && mkdir -p ~/.claude/skills && cp -r scientific-agent-skills/skills/database-lookup ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal (Terminal on macOS, Git Bash or WSL on Windows).
- Create the skills folder if needed:
mkdir -p ~/.claude/skills - Clone the repository:
git clone https://github.com/K-Dense-AI/scientific-agent-skills.git - Copy just this skill:
cp -r scientific-agent-skills/skills/database-lookup ~/.claude/skills/ - Verify the reference docs came along:
ls ~/.claude/skills/database-lookup/references | head— these files hold the endpoints and pagination rules. - Optional: for key-gated sources (FRED, NCBI, OpenFDA, Materials Project…), register for a free key and set it as an environment variable or in
~/.claude/.env, e.g.FRED_API_KEY=.... - Restart Claude Code and ask with a named source, e.g. "Look up aspirin in PubChem and cite the endpoint" — the skill triggers automatically.