Validate run_scientific_agent + analyze_dataset on local + Improv + UCAR (#28)#41
Merged
Merged
Conversation
Closes #28 by exercising both analyze_dataset and run_scientific_agent against six scenarios: - analyze_dataset: local healpix:2, local synthetic UGRID + data, Improv remote, UCAR remote - run_scientific_agent: local healpix:3, local synthetic UGRID + data (covers all four Analyze/Plan/Execute/Verify stages) All six scenarios pass. Top-level provenance reads venue=hpc on remote runs; per-stage inner provenance carries the full hpc:<endpoint_id>. The harness lives at scripts/validate_orchestrators.py and reads endpoints from config.yaml, so it's reproducible by anyone with the same endpoint UUIDs configured.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Closes #28. Adds a reproducible end-to-end validation harness for both the deterministic (
analyze_dataset) and autonomous (run_scientific_agent) orchestrators, and confirms they work across local + Improv + UCAR.scripts/validate_orchestrators.pyexercises six scenarios and reports a per-scenario pass/fail summary:analyze_dataset— local healpix:2 (no data)analyze_dataset— local synthetic UGRID + datarun_scientific_agent— local healpix:3run_scientific_agent— local synthetic UGRID + dataanalyze_dataset— REMOTE healpix:2 on Improvanalyze_dataset— REMOTE healpix:2 on UCARFor the remote scenarios, top-level provenance reads
venue=hpcand the per-stage inner provenance carries the fullhpc:<endpoint_id>so the actual execution venue is verifiable end-to-end. Forrun_scientific_agent, all four stages (Analyze → Plan → Execute → Verify) appear inreasoning_trace.Acceptance criteria from #28
run_scientific_agentcompletes all 4 stages on a local meshanalyze_dataset/orchestrators round-trip on Improv with the HPC endpointanalyze_datasetrecords per-stage warnings and continues;run_scientific_agentfalls through to the next stage on failureTest plan
uv run python scripts/validate_orchestrators.py→ 6/6 PASSuv run pre-commit run --all-files→ cleanuv run pytest tests/ --ignore=tests/test_remote_agent.py→ 247 passedFiles
scripts/validate_orchestrators.py— new harness (readsconfig.yaml)CHANGELOG.md— Unreleased entry under Fixed