Pipeline Export

Publication-grade research rarely lives in a single web UI — it lives in Nextflow, Snakemake, CWL, or WDL pipelines on a lab's HPC or in nf-core / Galaxy. Pipeline Export lets you take any BLAST search you ran from the dashboard and drop it into one of those pipelines, re-submitting the exact same parameter set through the dashboard API — without falling back to the elastic-blast CLI on a personal machine.

What you get

On a job's result page, the header carries a Pipeline export ▾ menu with four formats:

Format	Downloaded file	Run with
Nextflow	main.nf	nextflow run main.nf --query my.fasta
Snakemake	Snakefile	snakemake -j1 --config query=my.fasta
CWL	blast_submit.cwl	any CWL runner (e.g. cwltool blast_submit.cwl --query my.fasta)
WDL	blast_submit.wdl	any WDL runner (e.g. Cromwell / miniwdl)

Each module pins the source job's program, database, and algorithm options (E-value, word size, max_target_seqs, low-complexity filtering, taxonomy include/exclude, sharding mode, effective search space, resource profile). The query FASTA is the only runtime input — so the same module can be re-run against any number of query files.

How it works

The generated module makes a single authenticated POST /api/blast/jobs call with the pinned body, merging in the query FASTA you supply at run time. Nothing else is embedded:

• No storage URL and no SAS token are written into the file — the data plane stays private behind the dashboard API.
• No bearer token is baked in. The module reads credentials from the environment at run time (see below), so the file is safe to commit to a pipeline repo.
• idempotency_key is deliberately not pinned — every pipeline run is a fresh job, not a replay of the source job.

Runtime environment

The module reads three environment variables when it runs:

Variable	Meaning
ELB_BASE_URL	Base URL of your dashboard, e.g. https://ca-elb-dashboard.<region>.azurecontainerapps.io
ELB_TOKEN	A valid bearer token for the dashboard API (the same identity the browser uses)
ELB_QUERY_FASTA	Path to the query FASTA (the runners above set this for you from --query)

export ELB_BASE_URL="https://ca-elb-dashboard.<region>.azurecontainerapps.io"
export ELB_TOKEN="<your dashboard bearer token>"
nextflow run main.nf --query my.fasta

The call returns the dashboard's standard submit envelope (the new job's job_id and status URLs), which the pipeline step prints to stdout. Poll GET /api/blast/jobs/{job_id} (or open the dashboard) to follow the new run — its results land in the same workspace as any other dashboard search.

End-to-end loop

flowchart LR A[Dashboard search] -->|Pipeline export ▾| B[main.nf / Snakefile / .cwl / .wdl] B -->|drop into pipeline| C[Nextflow / Snakemake / CWL / WDL runner] C -->|POST /api/blast/jobs<br/>with your query FASTA| D[New BLAST job] D -->|results in the same workspace| A

1. Run a search from New Search, tune the parameters until you are happy with the result.
2. On the result page, open Pipeline export ▾ and download the module for your pipeline manager.
3. Commit the module to your pipeline repo and wire the query FASTA as its input.
4. Each pipeline run re-submits the search through the dashboard API; results appear back in the dashboard workspace, fully tracked and reproducible.

Notes and limits

• The export pins the parameters the dashboard recorded for the source job. If the job has no recorded database (an incomplete or external record), the export returns a clear 422 and the menu surfaces "Pipeline export is not available for this search yet".
• The generated module is intentionally dependency-light — it uses only the Python standard library to make the HTTPS call, so it runs anywhere python3 is available.
• A typed Python CLI (elb-dashboard-cli) generated from the OpenAPI spec is tracked separately and is not part of this export; this page covers the workflow-manager module files only.