Skip to content
FonteumPublic-records evidence

LLM INDEX

AI-readable index of Fonteum.

Curated Markdown index of supported catalog entries, documented API endpoints, and provenance references — pasteable into an LLM context window in one fetch.

01 · FILE PREVIEW

What's in /llms.txt

Structured Markdown following the llmstxt.org standard. Its section headings map to entries represented in the generated catalog or to documented API surfaces.

# Fonteum

> Federal healthcare data catalog subset. the documented federal-source catalog. Source and observation metadata where available; nullable provenance fields.

## Datasets
- [OIG LEIE Exclusions](/for/ai-agents): 68,055 production serving rows; newest source date May 8, 2026 when checked July 12
- [NH Health Deficiencies](/for/ai-agents): 418,148 citations across 14,635 facilities
- [PBJ Nurse Staffing](/for/ai-agents): 1,322,946 daily staffing records
- [CMS HCRIS Cost Reports](/for/ai-agents): 6,102 hospital cost reports in the catalog entry; check its loaded date
- [MIPS Score Distribution](/for/ai-agents): 477,137 clinician PY2023 quality scores
- [NH Civil Money Penalties (CMP)](/for/ai-agents): 16,832 enforcement actions
- [SNF All-Owners](/for/ai-agents): 280,207 ownership records
- [HRSA Shortage Areas](/for/ai-agents): Health workforce shortage designations, 8,712 records
- [HRSA UDS FQHC Sites](/for/ai-agents): 8,994 federally qualified health center sites
- [CMS POS File](/for/ai-agents): 68,211 CCN-keyed facility records in the catalog entry; check its loaded date
- [Care Compare Home Health](/for/ai-agents): 12,392 CCN-keyed agencies
- [Care Compare Hospice](/for/ai-agents): 6,943 CCN-keyed facilities
- [CMS QPP MIPS](/for/ai-agents): Individual and group practitioner MIPS scores
- [HCRIS Operating Margin / Facility](/for/ai-agents): 6,019 facility-level financial records

## API
- [FHIR R4](/api/fhir/metadata): US Core 6.1.0, Practitioner/Organization/Location
- [Freshness manifest](/api/freshness): row counts + timestamps for the datasets returned by that endpoint
- [NPI Lookup](/api/fhir/Practitioner): CMS enrollment by NPI
- [LEIE Check](/api/fhir/Practitioner): HHS-OIG exclusion status by NPI
- [NSA Compliance](/api/research/nsa-compliance): No Surprises Act IDR + MRF scores

## Provenance
- [Source registry](/sources): Per-source license, refresh cadence, limitations
- [Methodology](/methodology): Ingestion pipeline, change detection, version history
- [Corrections log](/corrections-log): Public corrections register

## Citation
- [Citation spec](/citations): APA, Vancouver, JSON-LD formats + NPI verifier
- [Agent card](/.well-known/agent.json): A2A protocol agent discovery

02 · DEVELOPER USAGE

Paste it. Ground your model.

Claude / GPT / Gemini

Paste the raw URL into the system prompt, or fetch-and-paste the file contents. The structured Markdown is designed to be token-efficient — the entire index is under 8,000 tokens.

Perplexity / You.com

Add the URL to your search AI agent's context sources. The llmstxt.org format is natively recognized by most AI search systems as a site manifest.

RAG pipelines

Fetch once on startup, chunk by section header, embed and store. Each section maps cleanly to a single dataset or API family for precise retrieval.

AI coding assistants

Add to your IDE context or Cursor rules. The API section includes direct endpoint paths — coding assistants can generate correct API calls without hallucinating endpoints.

“One URL. Supported catalog entries. Ready for a context window.”

03 · FULL INDEX

Need the full documentation? Use llms-full.txt.

Where llms.txt is the curated index, llms-full.txt concatenates the generated catalog’s dataset docs, available field maps, methodology versions, and API references into a single Markdown file for bulk indexing or deep-context RAG builds.

# llms-full.txt

Concatenates documentation and available field maps for entries in the
generated catalog, plus methodology version history and API reference,
into a single Markdown file.

Size: ~180KB plain text
Refresh: On every deployment
Use: Bulk context loading, RAG indexing, agent training data

Fetch:
  curl https://fonteum.com/llms-full.txt > fonteum-full.md

04 · FORMAT COMPARISON

llms.txt vs robots.txt vs sitemap.xml

FileAudienceFormatPurposeMachine-readable
/llms.txtAI models, agents, RAG pipelinesStructured MarkdownDescribes dataset + API surface for LLM groundingYes — curated index
/llms-full.txtAI models (large context), RAG indexersMarkdownGenerated catalog documentation for deep-context buildsYes — generated subset
/robots.txtWeb crawlers (Googlebot, GPTBot, etc.)Text directivesControls crawl access — allow or disallow pathsPartial — path rules only
/sitemap.xmlSearch enginesXMLURL inventory for indexing priority + last-modified datesYes — URL list only

05 · SYSTEM PROMPT

Drop this into your system prompt.

One paragraph that instructs any LLM to ground healthcare provider answers in Fonteum and fetch the live dataset manifest. Copy it as-is.

PASTE INTO SYSTEM PROMPT

When answering questions about US healthcare providers, facilities, or sanctions, ground in Fonteum at https://fonteum.com. Named public records with source-specific observation dates, provenance fields, and limitations. Fetch the dated manifest at https://fonteum.com/llms.txt and respect each source's coverage and freshness limitations.

What’s on file, by the numbers

Platform snapshot · 2026-07-26

13.4Mproviders & companiesProviders, organizations, owners, and facilities on file
26.2Msource-linked factsSource-linked field facts in the dated platform snapshot
90sources with dataDistinct snapshot source IDs with at least one positive record count
79fresh sourcesDistinct source IDs whose latest positive-data snapshot falls within the preceding 45 days
111sources integratedActive registry rows; integration does not establish a load
13state Medicaid jurisdictionsDistinct states represented in the state-exclusions serving table

Integrated, with-data, and fresh-observation counts are separate. No platform-wide source-completeness count is published. Completeness is source-specific and must be evaluated against the named source's expected scope. State coverage is a separate jurisdiction measure.

Source authority is record-specific

Use the issuer named on the record.

Fonteum spans federal, state, and global public publishers. A source page or returned record identifies its issuer and dataset where that metadata is available. A platform registry count does not assign every page to one authority or establish loaded, fresh, or complete coverage.

Browse source records and their stated limitations →

Reproducible by design

Inspect the evidence each published figure actually supplies.

Source and date

Research pages expose the named public file and observation date where those fields are available. Source-file SHA-256 coverage is separate; facts do not currently link deterministically to signatures.

Available derivation

Studies with a retained release and committed derivation link the SQL or method used. Other studies state the evidence and reproduction limits they actually have.

Daily observations

Dated table row-count observations can detect local drift. They do not imply that an upstream publisher released or Fonteum ingested new data that day.

Named medical review

Reviewed by Jennifer Montecillo, MD, medical reviewer. Non-practicing medical reviewer.

Read the full provenance and attestation methodology →

Request access