| name | repo-quality-gates |
| description | Run the test, lint, and typecheck commands for the repository, then summarize failures and next actions for coding agents. |
Repo Quality Gates
Use this skill when you need to validate a change or interpret failing automation for this repository.
Default validation commands
pip install langchain langchain-core langchain-openai langchain_experimental langgraphpython -m pytest test_validate_run.py -qpython validate_run.py --latest --log-path notebook_run_log.txt --window 180python validate_artifact_quality.py --latestpython -m pytest tests/flake8 .
IDD completion gates
For notebook-completion work, the current W14 baseline is IDD_run_run_default_id-20260504-1338-b3079aea. A clean proof means:
validate_run.py scores 12/12.validate_artifact_quality.py scores 9/9.- The log contains native structured-output markers and 3/3 visualization fan-in.
- The log has zero recovery/final-hop/native-failure/path-normalization warnings.
- Root
final_report.html, final_report.md, and final_report.pdf exist and embed resolving visualizations.
Workflow
- Run the smallest relevant validation subset first.
- Group failures by root cause instead of by raw output order.
- Call out missing tooling or environment issues separately from code regressions.
- Do not relax or delete tests just to produce a green run.
