bv-registry-staging

Auto-generated draft manifests for bv.

Do not depend on this repo directly. Manifests here are auto-generated from Bioconda by bv-ingest and may be incomplete. Reviewed manifests are promoted to bv-registry.

Layout

tools/
  <tool-id>/
    <version>.toml    # draft manifest; may lack typed I/O

Reviewer workflow

1. A nightly job opens one PR per tool to this repo (ingest/<tool>-<version> branch).
2. Open the PR, read the upstream README, fill in [[tool.inputs]] and [[tool.outputs]].
3. Verify [tool.entrypoint] is correct.
4. Merge the PR.
5. Run bv-ingest promote <tool> <version> --staging-dir <local-clone> to open a PR against bv-registry.

A review should take under 5 minutes. If a tool's I/O is genuinely unclear, leave a comment on the PR.

Available types

See bv-types/types.toml for the full vocabulary.

Common ones:

Type	Description
fasta	FASTA sequence file (fasta[protein], fasta[nucleotide])
fastq	FASTQ with quality scores
bam	Binary Alignment Map
sam	Sequence Alignment/Map (tabular)
vcf	Variant Call Format
pdb	Protein Data Bank structure
blast_db	BLAST database directory
blast_tab	BLAST tabular output (outfmt 6)
hmm_profile	HMMER profile
dir	Generic output directory

What bv-ingest fills in automatically

• tool.id, tool.version, tool.description, tool.license, tool.homepage
• tool.image.reference and tool.image.digest from quay.io/biocontainers
• tool.entrypoint.command (best-guess from recipe test commands)
• tool.hardware defaults (1 CPU, 2 GB RAM)

What you fill in

• [[tool.inputs]] - typed input ports
• [[tool.outputs]] - typed output ports
• Adjust [tool.hardware] if the tool has known requirements
• Correct [tool.entrypoint] if the auto-detected command is wrong