# Analyze SRA and GEO Genomics Metadata

Researchers often need to connect SRA, GEO, BioProject, and publication records before analysis. This skill guides Claude, Codex, or Claude Code through SRAgent commands for accession conversion, metadata inspection, and paper retrieval.

## Install

```bash
npx skillstore add arcinstitute/sragent
```

## Metadata

- - Slug: arcinstitute-sragent
- - Version: 1.0.0
- - Author: ArcInstitute
- - GitHub username: ArcInstitute
- - License: MIT
- - Repository: https://github.com/ArcInstitute/SRAgent/tree/main/claude-skill/
- - Ref: main
- - Supported tools: Claude, Codex, Claude Code
- - Risk level: medium
- - Risk factors: external\_commands, network, filesystem, env\_access
- - Quality score: 50
- - Quality tier: warning
- - Public page: https://skillstore.pages.dev/skills/arcinstitute-sragent
- - Manifest: https://skillstore.pages.dev/api/skills/arcinstitute-sragent/manifest

## Capabilities

- Converts GEO, SRA, BioProject, experiment, and run accession formats using SRAgent command examples.
- Guides metadata extraction for organism, platform, library layout, tissue, and single-cell technology fields.
- Supports paper lookup and open-access PDF download workflows linked to SRA accessions.
- Provides batch workflows for CSV files containing multiple SRA or GEO accessions.
- Documents BigQuery-backed analysis paths and Entrez-only fallback behavior.
- Includes reference tables for common accession patterns and sequencing metadata fields.

## Use Cases

- Convert accession identifiers: Convert a GEO series or BioProject accession into SRA experiment and run accessions for downstream analysis.
- Audit sequencing metadata: Check whether public experiments match required organism, platform, paired-end, or single-cell criteria.
- Collect dataset publications: Find linked PubMed records, DOIs, and available full-text papers for a set of SRA accessions.

## Prompt Templates

### Convert one accession

```
Use SRAgent to convert GSE121737 into SRA experiment accessions. Summarize the result in a short table.
```

### Check experiment metadata

```
Use SRAgent to check SRX4967527 for organism, platform, library layout, and whether it is single-cell RNA-seq.
```

### Download linked papers

```
Use SRAgent to find publications for PRJNA498286 and download available open-access papers into a dedicated output folder.
```

### Run a batch curation workflow

```
Given my CSV of SRA accessions, use SRAgent to add organism, sequencing platform, single-cell status, DOI, and download status columns.
```

## Limitations

- Requires a separate SRAgent installation before commands can run.
- Needs API keys or credentials for some model, NCBI, CORE, and BigQuery features.
- Paper downloads depend on open-access availability and source rate limits.
- Metadata completeness depends on public database records and may vary by accession.

## Best Practices

- Validate accession formats before running shell commands or batch workflows.
- Use redacted environment checks instead of printing API key values or .env contents.
- Start with Entrez conversion, then run deeper metadata or paper workflows on confirmed accessions.

## Anti Patterns

- Do not paste raw API keys, service account files, or .env contents into the chat.
- Do not run paper downloads at high concurrency when source services are rate limited.
- Do not assume every public accession has complete metadata or an open-access publication.

## Security Audit

- - Safe to publish: true
- - Audited at: 2026-06-28T10:25:52.864\+00:00
- - Summary: Static findings are mostly expected for a research workflow skill that runs the SRAgent CLI, queries public genomics services, and downloads papers. No prompt injection, credential exfiltration, or malicious intent was found, but troubleshooting guidance can expose API keys or .env contents and should be treated as a publication warning.

## Stats

- - Views: 524
- - Downloads: 6
- - Favorites: 0
- - Popularity score: 0