3.5 Run the Toy Dataset in Five Minutes

Use the 40-sample abundance-table example in BioLine/data/toy_data/ to check the basic path from input inspection to downstream result records. This uses an exact version of an existing BioLine downstream Tool and does not require a historical image to be preinstalled.

Example request to OpenClaw:

“Use the toy data for a downstream statistical analysis. First inspect the abundance table and metadata, then discover a suitable BioLine Tool and show its exact version and resource requirements. Wait for my confirmation before running it, and record the script, command, logs, and result location.”



After confirmation, OpenClaw checks the Tool summary/config. The summary must declare analysis_lane: downstream and fixed_command: false, and the config must contain no fixed analysis command. OpenClaw then pulls the exact NAME:VERSION on demand, generates and saves a script for this dataset, and runs it in an isolated environment with networking disabled.

Afterward, review the exact Tool version, summary/config snapshot or hash, pull identity, script, actual command, logs, exit status, and outputs recorded for the task. One successful run does not validate a different dataset or Tool version.