trove

Reference tools for underused public healthcare data. Open-source parsers, Parquet/JSON bundles, and Claude Code skills for FDA novel drug approvals (NMEs) and U.S. nonprofit hospital reporting (CMS HCRIS Worksheet S-10 + IRS Form 990 Schedule H). Live at troveproject.com — MIT-licensed.

Try it

• troveproject.com/drugs/ — search 218 FDA novel drug approvals from 2021–2024 (CDER NMEs + CBER cell & gene therapies); click any record for its full approval package
• troveproject.com/hospitals/ — look up 1,295 nonprofit U.S. hospital systems; HCRIS and 990 Schedule H side-by-side
• troveproject.com/skills/ — install the trove Claude Code plugin (fda-analyst + hcris-analyst)
• troveproject.com/docs/ — plain-language explainers: what is HCRIS, what is IRS 990 Schedule H, what is an FDA novel drug approval

What this is

trove builds open-source lookup tools, parsers, and Claude Code skills on top of public-domain healthcare datasets that are widely cited but rarely usable in their raw form. CMS publishes 100,000-row long-skinny CSVs. The IRS publishes 990s as XML in bulk ZIPs. The FDA scatters drug-approval reviews across hundreds of PDF directories. trove does the parsing, joining, and packaging so the data is browsable and queryable.

Two areas live in v1:

/drugs — FDA drug approvals

Look up any FDA novel drug approval from 2021–2024 (218 drugs). v0.2 ingests both FDA regulatory centers: CDER (192 NMEs and novel antibody BLAs from the annual Novel Drug Approvals lists) plus CBER (26 cell and gene therapy approvals — Lenmeldy, Casgevy, Lyfgenia, Beqvez, Hemgenix, Roctavian, Adstiladrin, etc.). Each row carries application number, sponsor, approval date, indication, a regulatory_center column, and a deep link to either the drugs@FDA application overview (CDER) or the per-product FDA page (CBER) where every approval-package document lives. Companion Claude skill fda-analyst reads those PDFs at query time.

Sources: FDA's annual Novel Drug Approvals curated lists (CDER); Approved Cellular and Gene Therapy Products page (CBER); drugs@FDA database. US government work, public domain.

/hospitals — hospital reporting (CMS + IRS)

A lookup tool for related charity-care reporting fields that nonprofit U.S. hospitals file with two different regulators: CMS (Worksheet S-10 of the Medicare Cost Report) and the IRS (Form 990 Schedule H Part I line 7a). Both lines are intended to capture the cost of care provided to patients who couldn't pay, but the rules and scope diverge enough that the two numbers can legitimately differ. Search any of 1,295 nonprofit hospital systems for tax year 2022 and see the two filings side-by-side, with period-alignment context, a home-county Social Vulnerability Index proxy, and a deep link to the actual 990 on ProPublica.

For tax year 2022, the funnel: 1,334 systems matched at the EIN level → 1,295 computable → 372 period-aligned within 1 month → 228 also material (both filings ≥ $500K). Among those 228, the median absolute proportional difference is 25%.

Full data: artifacts/community_benefit_gap_2022.csv · Method: artifacts/community_benefit_gap_2022_summary.md.

Reproduce

git clone https://github.com/cbetz/trove
cd trove && uv sync --all-packages
# /hospitals dataset:
uv run python scripts/build_gap_dataset.py
# /drugs dataset:
uv run python scripts/build_fda_index.py

What's in the box

Hospital reporting area:

• hcris — CMS Medicare Cost Reports (form 2552-10) parser with a 44-variable semantic dictionary. Raw HCRIS ships as headerless long-skinny CSVs; this turns them into tidy DataFrames and partitioned Parquet.
• form990 — IRS Form 990 Schedule H parser. Bulk-XML download, index reader, and 19 fields per filing including Part I 7a–k community benefit amounts and Part III bad debt.
• crosswalk — CCN ↔ EIN crosswalk (3,523 hospitals, 2,385 EINs). Bundled from Community Benefit Insight.
• analytics — composed queries. community_benefit_gap() is the cross-form charity-care difference primitive.
• sdoh — Social Determinants of Health enrichments. Ships CDC Social Vulnerability Index 2022 county-level (public domain — included in the public bundles) and UW Area Deprivation Index county aggregation (local-only — UW's terms are non-sublicensable).

FDA drug approvals area:

• fda_sba — Scrapes FDA's novel-approval sources from both CDER (annual Novel Drug Approvals pages) and CBER (Approved Cellular and Gene Therapy Products page). Extracts application number / drug name / active ingredient / approval date / sponsor / indication / regulatory_center, and emits links to the approval-package PDFs. v0.2 covers 2021–2024 (218 drugs: 192 CDER + 26 CBER).

Site: static, deployed to Vercel at troveproject.com.

Install the skills

trove is packaged as a Claude Code plugin (.claude-plugin/plugin.json) containing both skills. Two install paths:

Plugin install (recommended once published to the official marketplace):

/plugin marketplace add cbetz/trove
/plugin install trove@trove

Filesystem install (works today):

git clone https://github.com/cbetz/trove
cp -r trove/skills/hcris-analyst ~/.claude/skills/
cp -r trove/skills/fda-analyst ~/.claude/skills/

Either way, restart Claude Code and both skills are loaded. Each skill's description field tells Claude when to invoke — you don't call them by name.

• skills/hcris-analyst/ — natural-language queries over hospital reporting (HCRIS Worksheet S-10 + IRS 990 Schedule H + CCN↔EIN crosswalk + CDC SVI). Profile lookups, peer context, glossary, and cross-form difference context.
• skills/fda-analyst/ — questions about specific FDA drug approvals. Reads approval-package PDFs at query time.

For per-project install (.claude/skills/ inside the repo), Claude.ai upload, or authoring a new skill, see skills/README.md.

Dev setup

uv sync --all-packages
uv run pytest
uv run ruff check

Layout

packages/         Python libraries (hcris, form990, crosswalk, analytics, sdoh, fda_sba)
skills/           Claude skill bundles
web/              Static site at troveproject.com
pipelines/        ETL orchestration — TBD
notebooks/        Exploratory work, not shipped
docs/             mkdocs-material site
artifacts/        Committable, GitHub-viewable result tables
scripts/          Build + demo scripts

Status

• M6 — Multi-area site: trove now has two live areas at troveproject.com (/drugs and /hospitals), each with its own search-first lookup page, methodology, and Claude skill. New fda_sba package indexes FDA Novel Drug Approvals 2021–2024 (~192 drugs); new fda-analyst Claude skill reads approval-package PDFs at query time. Adding more areas in the future is now a matter of adding a package + a page + a skill.
• M5.2 — CDC Social Vulnerability Index added to sdoh and wired through to the public bundles + the troveproject.com detail card. Each row now shows a "Service area (SVI)" signal: e.g. "81st national percentile — high social vulnerability in the system's home county". SVI is fully public-domain, so unlike ADI this ships in the public CSV / Parquet / JSON. ADI capability remains in the local pipeline for users with their own UW license.
• M5.1 — sdoh package added with county_adi_from_block_group(). ADI flows through the local pipeline only — UW's non-sublicensable terms mean it can't be in the public bundles.
• M5 — hcris-analyst Claude skill shipped. Natural-language queries against HCRIS + 990 + the matched hospital dataset, powered by DuckDB-over-HTTPS against three Parquet bundles served from troveproject.com (hcris_2023_wide.parquet, schedule_h_2022.parquet, community_benefit_gap_2022.parquet). Skill bundle at skills/hcris-analyst/ includes the SKILL.md and reference docs (field dictionary, peer-cohort definitions, Schedule H map, runnable example queries).
• M4.3 — Validation pass against IRS source XML (6 spot-checks, all exact); per-row alignment signals (HCRIS fiscal year end vs. 990 tax period end); default view filtered to aligned + material subset (228 systems); ProPublica deep links per row.
• M4.2 — TY2022 ingest expanded to 2024+2025+2026 IRS release years (late filers and amendments). Form990 parser now handles DEFLATE64 ZIPs (introduced by IRS in 2025 release).
• M4.1 — first published TY2022 matched hospital dataset from the 2024 release year alone (1,317 systems).
• M4 — crosswalk package + analytics.community_benefit_gap().
• M3 — form990 Schedule H parser, full TY2022 ingest.
• M2.5 — HCRIS semantic dictionary expanded to 44 variables.
• M2 — hcris.pivot_wide() and dictionary v1.
• M1 — hcris parser for Hospital 2552-10.

Per-package detail in packages/*/README.md.

License and citations

trove code is MIT-licensed. Underlying data sources have their own licensing and citation requirements:

• CMS HCRIS — US government work, public domain. No citation required; suggested phrasing: "CMS Healthcare Cost Report Information System (HCRIS), Hospital form 2552-10".
• IRS Form 990 e-file — US government work, public domain. Suggested: "IRS Tax-Exempt Organization Form 990 e-file (Schedule H)".
• CCN ↔ EIN crosswalk — derived from Community Benefit Insight (RTI International, funded by RWJF, frozen Dec 6 2024). Suggested: "Community Benefit Insight; RTI Press 10.3768/rtipress.2023.op.0080.2302".
• CDC Social Vulnerability Index — county-level 2022 release, US government work / public domain. Included in the public bundles as svi_overall_pct plus four sub-theme columns. Source: https://www.atsdr.cdc.gov/place-health/php/svi/index.html.
• Area Deprivation Index — UW's Neighborhood Atlas terms are non-sublicensable and forbid redistribution of the data, derived or otherwise, outside individual-and-employer "internal non-profit educational, research, and public health" use. trove does not include ADI in the public bundles for this reason. The sdoh package supports local pipelines: download the block-group CSV from https://www.neighborhoodatlas.medicine.wisc.edu/ (registration required) and the matched hospital dataset built locally will include county-aggregated ADI columns. Required citation when you use that local output anywhere: Kind AJH, Buckingham W. Making Neighborhood Disadvantage Metrics Accessible: The Neighborhood Atlas. N Engl J Med 2018;378:2456-2458; PMCID: PMC6051533. AND: University of Wisconsin School of Medicine and Public Health. {year} Area Deprivation Index {version}. Downloaded from https://www.neighborhoodatlas.medicine.wisc.edu/ {date}.