Repository navigation
Build an evidence-first astronomy research workbench - #1
Merged
Merged
Conversation
RyanSingh0
marked this pull request as ready for review
November 11, 2026 01:01
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The repository previously contained only the project plan. This adds a usable local workbench for inspecting NASA archive question evidence, comparing candidate result tables, and collecting consented scientific reviews.
The Python package includes 20 public development questions, 17 captured references with raw VOTables and hashes, 48 capture checks, unit-aware ordered/multiset comparison, bounded TAP capture, SQLite persistence, and an offline FastAPI/browser interface. Known direct-TAP TOP/ORDER ranking behaviour is quarantined and the ranking reference is independently checked. Documentation separates tested software from pending scientific approval and real-user validation.
Validation: 35 local tests pass; lint/format checks pass; wheel installation in a clean environment and evidence verification pass. Browser checks cover incorrect and reference results, saved reviews, search, responsive layouts, and evidence download. CI adds Windows/Linux checks and container smoke testing. No paid model calls or model accuracy claims.
Keep this draft while scientific review and the BU/LinkedIn pilot are pending. The application is local-only; production hosting, model experiments, and a held-out benchmark remain separate release gates.