# Esearch

> Use when searching NCBI Entrez databases (pubmed, gene, protein, nuccore, snp, geoprofiles) with query strings and field qualifiers to retrieve record UIDs for downstream processing.

- Skill: `vimalinx/esearch` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add vimalinx/esearch`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vimalinx/esearch/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: vimalinx (https://skillmd.com/u/vimalinx)
- Updated: 2026-08-19
- Page: https://skillmd.com/skills/vimalinx/esearch

---


# esearch

## Quick Start
- **Command:** `esearch -db <database> -query "<query string>"`
- **Local executable:** `/home/vimalinx/miniforge3/envs/bio/bin/esearch`
- **Full reference:** See [references/help.md](references/help.md)

## When To Use This Tool

- Start an Entrez Direct pipeline by searching for record IDs.
- Search PubMed, Gene, Protein, Nuccore, SRA, and related NCBI databases with field-qualified queries.
- Hand off matching IDs to `efetch`, `esummary`, `elink`, or `xtract`.
- Use this whenever you need database-specific search syntax, not a free-form web search.

## Common Patterns

```bash
# 1) PubMed literature search
esearch -db pubmed -query 'ebola virus[Title/Abstract] AND 2024[pdat]'
```

```bash
# 2) Gene search with field qualifiers
esearch -db gene -query 'TP53[gene] AND human[orgn]'
```

```bash
# 3) Search and immediately fetch document summaries
esearch -db assembly -query 'GCF_000001405.40[accn]' | efetch -format docsum
```

## Recommended Workflow

1. Pick the Entrez database first, because query fields and sort modes are database-specific.
2. Write the query with field tags whenever possible instead of relying on broad keywords.
3. Run `esearch`, then immediately pipe IDs into `efetch`, `esummary`, or `elink`.
4. Use `xtract` only after you know what XML structure the downstream command emits.

## Guardrails
- `-db` and `-query` are both required.
- Wildcards and unqualified terms can explode result counts; narrow with fields like `[AUTH]`, `[GENE]`, or `[orgn]`.
- Sort options are database-specific, so do not assume the same `-sort` values work everywhere.
- `esearch` gives you IDs, not the final report; plan the next pipeline step before running it at scale.