Skip to content
Get packs

TileDB-VCF Genomic Variant Store

Lets an agent build, query and export indexed genomic variant (VCF/BCF) cohort datasets with TileDB-VCF.

Source Download skill
Repository
K-Dense-AI/scientific-agent-skills
Stars
55,780
Licence
MIT
Files
2 files

What it does

When installed, the agent gains detailed instructions for TileDB-VCF, a tool that stores many single-sample genomic variant files (VCF/BCF) in one indexed dataset so they can be queried by genomic region, sample, or attribute. The skill tells the agent how to install the right native software, check input files (one sample per file, sorted, indexed), ingest samples incrementally, run region queries with correct coordinate conventions, stream large results, and export subsets. It also warns about common pitfalls such as mixed genome assemblies and memory limits.

Great for

  • Bioinformatics teams storing and querying cohort-scale variant calls
  • Developers building genomics pipelines that need fast region or sample lookups
  • Researchers who want an agent to help set up and check a TileDB-VCF workflow

Skip it if

  • Joint variant calling, association testing or normalization, which the skill explicitly says TileDB-VCF does not do; using it for these would produce wrong analyses
  • Clinical or patient-facing decisions without expert validation, since the skill is based on synthetic test data and a single macOS ARM64 environment
  • Users without bioinformatics background, because mistakes in assembly, coordinates or sample identity silently produce incorrect results

Try asking

  • “Set up TileDB-VCF on my Mac and check that the Python import works”
  • “Ingest these sorted, indexed single-sample VCF files into a new cohort dataset”
  • “Query variants in chr1:1-100000 for samples S1 and S2 and export the results to a DataFrame”
  • “Explain why my deletion does not show up at the position I queried”
Step-by-step guide: how to use TileDB-VCF Genomic Variant StoreInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    tiledbvcf.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip tiledbvcf.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip tiledbvcf.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Give the agent your reference assembly, contig naming and sample IDs up front so it can verify inputs before ingesting.
  • Test ingestion and export on a small known interval before scaling to a full cohort.
  • Review any conda install commands and remote storage credentials yourself before running them, and keep your token out of shared prompts.

What's inside

  • references/verification.md
  • SKILL.md

Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.