fix: normalize supporting study provenance

[SkyeAv]

Moves supporting-study output to Biolink 4.4.4 Study metadata, removes synthesized # identities and row names, and preserves provenance with scoped row:<N> results.

Supporting-study provenance

• Identity: Study.id uses the publication CURIE or config stem; Study.name uses the worksheet/source filename.
• Metadata: Canonical study_* fields are emitted on Study; legacy aliases remain accepted.
• Results: StudyResult.id is row:<N> with no generated name. Contentless unpublished studies are omitted.
• Compatibility: Duplicate study-size aliases are dropped to prevent provenance leakage.

Biolink 4.4.4

• Derives field ownership and enums from the installed model.
• Emits effect_size as numeric and significance qualifiers as bare enum tokens.
• Adds explicit openpyxl test dependency and updates uv.lock.

Documentation

Updated README, CLI/configuration docs, tutorials, examples, agent prompts, tests, and CHANGELOG.md.

Testing

• uv run pytest tests/ -q --no-header -p no:cacheprovider → 1023 passed, 38 skipped
• uv run pre-commit run --all-files → passed
• uv run pre-commit run --all-files --hook-stage pre-push → passed, including Pyright and Clippy
• uv lock --check → passed
• cargo fmt --manifest-path rust/Cargo.toml --check → passed

cargo test remains blocked locally by a stale Nix linker-wrapper path; no Rust files changed.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 0afedbe6-58e9-4cb6-90ef-ed113719b314

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}