product stringclasses 1
value | track stringclasses 1
value | registry_id stringclasses 1
value | endpoint stringclasses 1
value | server_version stringclasses 1
value | observed_at timestamp[s]date 2026-07-30 03:24:30 2026-07-30 03:24:30 | name stringlengths 14 24 | title stringlengths 21 35 | description stringlengths 168 572 | input_schema unknown | output_schema dict | annotations dict |
|---|---|---|---|---|---|---|---|---|---|---|---|
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | device_risk_lookup | Review Device Category Risk Signals | Review a medical-device category's public FDA signals by three-letter product code (e.g. FRN = infusion pump). Returns recalls, MAUDE adverse-event trend, warning-letter matches, a normalized category signal, its driver contributions, and interpretation limits. It does not predict enforcement against a firm. | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"product_code": {
"type": "string",
"minLength": 3,
"maxLength": 3,
"description": "FDA product code, e.g. FRN"
}
},
"required": [
"product_code"
]
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Review Device Category Risk Signals",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | firm_compliance_history | Review Firm Compliance History | Build a recent, source-bounded FDA public-record timeline for a device firm: matched recalls, warning letters, and Form 483 citations where exact FEI numbers are available. Product codes are discovered from Constat's AI/ML-device corpus or may be supplied explicitly. Returns attribution and coverage limits with the rec... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"firm_name": {
"type": "string",
"minLength": 2,
"maxLength": 160,
"description": "FDA applicant or company name, e.g. 'Medtronic'"
},
"product_codes": {
"description": "Optional FDA... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Review Firm Compliance History",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | watchlist_diff | Poll Watchlist Changes | Return machine-generated FDA public-record changes detected for monitored product codes since a caller-supplied date, plus each code's latest category snapshot and postmarket coverage. Defaults to Constat Radar's five-code watchlist and the last seven days. Analyst verdict text and internal review status are excluded; ... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"product_codes": {
"description": "FDA product codes to poll. Omit for the default Radar watchlist.",
"maxItems": 20,
"type": "array",
"items": {
"type": "string",
"minLength": 3,
... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Poll Watchlist Changes",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | device_evidence_lookup | Look Up Device Evidence | Look up the structured premarket evidence FDA accepted for a specific AI/ML-enabled device by 510(k) number (e.g. K252148). Returns parsed summary fields — validation study design, sample sizes, endpoints, reported performance, predicate chain, PCCP — each with a verbatim source quote and page. Null means the summary d... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"k_number": {
"type": "string",
"pattern": "^(K|DEN)\\d{6}$",
"description": "510(k) or De Novo number, e.g. K252148 or DEN180001"
}
},
"required": [
"k_number"
]
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Look Up Device Evidence",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | evidence_search | Search Device Evidence | Find AI/ML device clearances by filter — product code, panel, applicant, and whether the submission reported clinical data, any sensitivity metric, or a PCCP. Answers 'what evidence did FDA accept for devices like mine'. Returns matching records with their parsed evidence. Presence flags are descriptive: 'reports a sen... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"product_code": {
"description": "FDA product code, e.g. QAS",
"type": "string",
"minLength": 3,
"maxLength": 3
},
"panel": {
"description": "Advisory panel, e.g. Radiology",
"... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{
"type": "object",
... | {
"title": "Search Device Evidence",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | predicate_chain | Trace Predicate Chain | Trace the predicate ancestry of a 510(k) device, with each cited predicate's age (how many years old the predicate was when the child cleared). Reveals how AI/ML devices chain to older predicates. | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"k_number": {
"type": "string",
"pattern": "^(K|DEN)\\d{6}$",
"description": "510(k) or De Novo number, e.g. K252148 or DEN180001"
},
"depth": {
"description": "Max ancestry depth (default... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{
"type": "object",
... | {
"title": "Trace Predicate Chain",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | evidence_cohort_stats | Summarize Evidence Cohort | Reporting-rate stats across the parsed AI/ML corpus (optionally by panel). Each rate is a presence figure with its denominator — 'reported in X of Y audited devices' — never a pooled performance value. Excludes not-yet-parsed devices from every denominator and discloses the parse queue separately. Predicate age (median... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"panel": {
"description": "Advisory panel, e.g. Radiology; omit for all",
"type": "string"
}
}
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Summarize Evidence Cohort",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | device_postmarket_lookup | Look Up Device Postmarket Signals | Post-clearance intelligence for one AI/ML device by 510(k) number: its product code's recalls, MAUDE adverse-event level and trend, warning-letter and 483 matches for the applicant, plus per-device drift signals (adverse-event inflection, re-clearances of the same device line, software-recall patterns, predicate-cohort... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"k_number": {
"type": "string",
"pattern": "^(K|DEN)\\d{6}$",
"description": "510(k) or De Novo number, e.g. K252148 or DEN180001"
}
},
"required": [
"k_number"
]
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Look Up Device Postmarket Signals",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | postmarket_search | Search Postmarket Signals | Find AI/ML devices by postmarket criteria — product code, panel, applicant, whether any drift signal exists, minimum recalls in 24 months, or a rising MAUDE trend. Returns per-device postmarket summaries with drift-signal counts. | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"product_code": {
"description": "FDA product code, e.g. QIH",
"type": "string",
"minLength": 3,
"maxLength": 3
},
"panel": {
"description": "Advisory panel, e.g. Radiology",
"... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{
"type": "object",
... | {
"title": "Search Postmarket Signals",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | cohort_postmarket_stats | Summarize Postmarket Cohort | Postmarket presence rates across the snapshotted AI/ML device cohort (optionally by panel): share with any recall in 24 months, with a rising MAUDE trend, with any drift signal, with a warning-letter match — every rate with its denominator inline, never pooled across devices. | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"panel": {
"description": "Advisory panel, e.g. Radiology; omit for all",
"type": "string"
}
}
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Summarize Postmarket Cohort",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | reimbursement_lookup | Look Up Reimbursement Pathways | Trace the clearance-to-payment pathway for an AI/ML device by FDA clearance number (K/DEN, e.g. DEN170073) OR bare CPT code (e.g. 75580). Returns every payment mechanism (NTAP add-on, Category I/III CPT + CMS rate, HCPCS, MAC LCD) with amounts, effective dates, and source links, plus any commercial/MAC payer coverage p... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"k_number": {
"description": "FDA clearance number, e.g. DEN170073 or K252148",
"type": "string",
"pattern": "^(K|DEN)\\d{6}$"
},
"cpt_code": {
"description": "Bare CPT code, e.g. 75580 or... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Look Up Reimbursement Pathways",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | reimbursement_search | Search Reimbursement Pathways | Find AI/ML device payment pathways by mechanism — e.g. 'devices that got NTAP', 'devices paid under a Category I CPT code', 'pathways with a known CMS dollar rate'. Filters: mechanism, CPT category, NTAP status, applicant. Returns pathways with amounts, effective dates, and sources. Use reimbursement_stats for the mech... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"mechanism": {
"description": "Payment mechanism",
"type": "string",
"enum": [
"ntap",
"cat_iii_apc",
"cat_i",
"hcpcs",
"lcd",
"none"
]
},
"... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{
"type": "object",
... | {
"title": "Search Reimbursement Pathways",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | reimbursement_stats | Summarize Reimbursement Cohort | Distribution of payment mechanisms across the AI/ML reimbursement corpus — pathway and distinct-device counts per mechanism (NTAP, Cat I, Cat III/APC, …) with the min/median/max dollar amounts for each. Deliberately never a single pooled 'reimbursement rate': NTAP add-on amounts and CMS rates are different measurements... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {}
} | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Summarize Reimbursement Cohort",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat | Radar | com.healthai/radar | https://constat.dev/api/mcp | 0.5.3 | 2026-07-30T03:24:30 | vehicle_risk_lookup | Look Up Vehicle Safety History | Look up NHTSA safety history for a vehicle by make, model, and model year. Returns recall campaigns and complaint statistics (crashes, fires, injuries, top components). | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"make": {
"type": "string",
"minLength": 2,
"description": "e.g. honda"
},
"model": {
"type": "string",
"minLength": 1,
"description": "e.g. civic"
},
"year": {
"... | {
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"ok",
"not_found",
"invalid_request",
"unavailable"
]
},
"data": {
"anyOf": [
{},
{
"type"... | {
"title": "Look Up Vehicle Safety History",
"readOnlyHint": false,
"destructiveHint": false,
"openWorldHint": false
} |
Constat MCP Tool Contracts
This dataset is a dated, machine-readable snapshot of the public tool contract
served by Constat MCP (com.healthai/radar). Radar is Constat's regulatory
and reimbursement intelligence track.
It contains tool names, descriptions, JSON input and output schemas, explicit side-effect annotations, endpoint metadata, and a validation report. It does not contain user prompts, user records, production analytics, private analyst notes, or Constat's regulatory evidence corpus.
The tools summarize source-linked public records and their limits. They do not make clinical decisions, certify device safety, predict enforcement, or act on behalf of FDA, CMS, NHTSA, or any payer.
Files
tools.jsonl: one public MCP tool descriptor per rowsnapshot.json: endpoint, version, observation time, and row countreport.json: annotation and output-schema completenesschecksums.sha256: integrity hashes for the generated files
Use the snapshot
Load the tool contracts directly with datasets:
from datasets import load_dataset
tools = load_dataset(
"healthai-hq/constat-mcp-tools",
split="train",
)
print(tools.column_names)
print(tools[0]["name"])
Useful applications include MCP client fixtures, tool-discovery experiments, schema regression tests, directory indexing, and comparing contract changes between dated releases. The rows describe the public interface; they are not a regulatory benchmark and do not reproduce the source records returned by the live service.
Snapshot method
The snapshot is generated from the production server's MCP tools/list
response with:
node distribution/huggingface/build-contract-dataset.mjs \
--input /path/to/tools-list-response.txt \
--output distribution/huggingface \
--observed-at 2026-07-29T00:00:00Z \
--endpoint https://constat.dev/api/mcp \
--registry-id com.healthai/radar \
--server-version 0.5.3
report.json records structural checks such as annotation and output-schema
coverage. Each published update is tied to the endpoint, server version, and
observation timestamp in snapshot.json; checksums make downloaded snapshots
independently comparable.
Limitations
- A tool-contract snapshot does not establish regulatory accuracy, security, or uptime.
- The live MCP server may be newer than this dated release.
- Tool descriptions and schemas are authored metadata, not FDA, CMS, NHTSA, or payer records.
- Consumers should use the production endpoint for current results and retain the snapshot only when reproducibility matters.
License and attribution
The metadata is Health AI-authored public interface documentation. license: other is intentionally conservative; this package does not relicense FDA,
CMS, NHTSA, payer, or other third-party records returned by the tools.
Product: https://constat.dev
MCP endpoint: https://constat.dev/api/mcp
Repository: https://github.com/healthai-hq/fda-risk-radar
Terms: https://constat.dev/terms
Support: hello@healthai.com
- Downloads last month
- 42