tessl install tessl/pypi-luigi@2.8.0Python workflow management framework for building complex pipelines of batch jobs with dependency resolution and task scheduling.
Agent Success
Agent success rate when using this tile
72%
Improvement
Agent success rate improvement when using this tile compared to baseline
1.31x
Baseline
Agent success rate without this tile
55%
{
"context": "Evaluates whether the pipeline runner relies on Luigi's supported entrypoints and scheduler coordination features to execute tasks, select between local and remote scheduling, and surface run results. Scoring emphasizes proper use of Luigi APIs rather than custom orchestration or generic concurrency code.",
"type": "weighted_checklist",
"checklist": [
{
"name": "Luigi entrypoint",
"description": "Pipeline is launched through Luigi's official entrypoints (e.g., luigi.build or luigi.run) with declared Task classes instead of custom scheduling logic.",
"max_score": 20
},
{
"name": "Local scheduler flag",
"description": "Local-only mode sets Luigi's local scheduler option (such as local_scheduler=True or the --local-scheduler flag) so tasks run without contacting a central scheduler.",
"max_score": 20
},
{
"name": "Remote scheduler config",
"description": "Remote mode wires scheduler host/port arguments into Luigi's scheduler parameters (e.g., scheduler_host/scheduler_port or --scheduler-host/--scheduler-port) before invoking the runner, and aborts cleanly on connection failures.",
"max_score": 20
},
{
"name": "Worker limit",
"description": "Requested worker counts are passed to Luigi's worker control (workers argument in luigi.build/luigi.run or equivalent) rather than custom threading, ensuring Luigi enforces the concurrency limit.",
"max_score": 15
},
{
"name": "Exit codes from result",
"description": "Process exit code derives from Luigi's run outcome (e.g., BuildResult.status or LuigiStatusCode) instead of ad-hoc checks, mapping success to 0 and failures to non-zero values.",
"max_score": 15
},
{
"name": "Run summary output",
"description": "Completion summary is generated from Luigi-provided run metadata (such as execution_summary or BuildResult.summary_dict) to report completed/failed/pending counts, not manually inferred.",
"max_score": 10
}
]
}