France Data MCP

professionnel_by_rpps

Read-onlyIdempotent

Retrieve a healthcare professional's complete record by RPPS/IDNPS identifier. Returns one entry per practice site with geolocation precision, falling back to live API when not found locally.

Instructions

Récupère la fiche complète d'un PS par identifiant national (rpps_id / IDNPS, 11 ou 12 chiffres — IDs émis depuis 2020 ont un préfixe "81" = 12 chars ; anciens IDs = 11 chars). Renvoie N entrées quand le PS exerce sur plusieurs sites (1 par site, chacun avec sa propre geo_precision — un même PS peut donc cumuler un site précis FINESS et un site au centroïde commune).

Chaque résultat géolocalisé porte geo_precision ∈ {"adresse", "etablissement_finess", "centroide_commune"} — lire ce champ pour évaluer la fiabilité des coords (précise BAN/FINESS au m près vs centroïde commune ~3 km, non discriminant intra-commune).

Fallback automatique sur l'API FHIR ANS live (gateway.api.esante.gouv.fr/fhir/v2) si non trouvé en base locale (snapshot mensuel J-30 max). Le champ source distingue db (base locale) de ans_fhir (live). include_freshness n'affecte que source: "db". Source : Annuaire Santé, Agence du Numérique en Santé (ANS) — Licence Ouverte v2.0

Input Schema

TableJSON Schema

Name	Required	Description	Default
`rpps_id`	Yes
`include_freshness`	No	Si true, ajoute un champ `data_freshness` au payload (dans `query_metadata` si présent, sinon à la racine) listant, par source (FINESS, Ameli, RPPS, CDS, IRIS), la dernière ingestion réussie (`last_success_at`, `staleness_days`) ET la dernière fois que la donnée a réellement changé (`last_data_change_at`, `data_age_days` — un run court-circuité « fichier amont identique » compte comme succès mais ne rajeunit pas la donnée ; c'est `data_age_days` qui dit l'âge réel de ce qui est servi). Opt-in pour ne pas alourdir les payloads par défaut. Cache 5min côté serveur — coût négligeable.

Output Schema

TableJSON Schema

Name	Required	Description
`key`	No	Clé recherchée (SIREN, num_finess, code INSEE, …).
`found`	Yes
`message`	No	Explication actionnable quand `found=false` (cause probable + remédiation).
`lookupStatus`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedv0.30.0Sep 7, 2026

changedInput schema / properties / include_freshness / description

Previous value: -"Si true, ajoute un champ `data_freshness` au payload (dans `query_metadata` si présent, sinon à la racine) listant la dernière ingestion réussie par source (FINESS, Ameli, RPPS, CDS) avec `staleness_days`. Opt-in pour ne pas alourdir les payloads par défaut. Cache 5min côté serveur — coût négligeable."New value: +"Si true, ajoute un champ `data_freshness` au payload (dans `query_metadata` si présent, sinon à la racine) listant, par source (FINESS, Ameli, RPPS, CDS, IRIS), la dernière ingestion réussie (`last_success_at`, `staleness_days`) ET la dernière fois que la donnée a réellement changé (`last_data_change_at`, `data_age_days` — un run court-circuité « fichier amont identique » compte comme succès mais ne rajeunit pas la donnée ; c'est `data_age_days` qui dit l'âge réel de ce qui est servi). Opt-in pour ne pas alourdir les payloads par défaut. Cache 5min côté serveur — coût négligeable."

Addedv0.20.0May 26, 2026
Removedv0.12.2May 20, 2026
Addedv0.10.3May 15, 2026
Removedv0.10.0May 15, 2026
Addedv0.9.3May 14, 2026
Removedv0.9.2May 14, 2026
First observedv0.7.6May 13, 2026

TDQS

A4.8/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint, openWorldHint, idempotentHint, and non-destructive behavior. The description adds significant behavioral context beyond annotations: it explains that a single PS may return N entries (one per practice site) with distinct `geo_precision` values, describes the fallback to the live ANS FHIR API, clarifies the `source` field semantics, and details how `include_freshness` affects only `source: 'db'`. This is rich, non-redundant disclosure.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is long but every sentence earns its place. It is front-loaded with the core purpose and ID format, then systematically covers multiplicity, geo_precision reliability, fallback behavior, source distinction, and freshness semantics. There is no redundancy or fluff; structure is logical and information-dense.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (multiple sites, geo_precision levels, live fallback, source differentiation, freshness semantics), the description covers all aspects an agent needs to call it correctly. It explains expected output multiplicity, field meaning, and edge cases (e.g., data age vs last success), and there is an output schema to further document return structure. Nothing essential is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is only 50% (one parameter has a schema description). The description compensates fully: it explains the `rpps_id` format (11/12 digits, prefix rules) beyond the schema's pattern, and clarifies that `include_freshness` only affects `source: 'db'`. This adds meaning beyond the structured schema and covers both parameters effectively.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the verb ('Récupère'), the resource ('la fiche complète d'un PS'), and the identifier (`rpps_id` / IDNPS). It specifies the exact ID format (11 or 12 digits, prefix '81' for post-2020 IDs) and distinguishes this tool from sibling tools by its by-ID lookup scope, making its purpose unambiguous and distinct from radius/name-based searches.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides clear context on when to use this tool (fetch by national ID) but does not explicitly mention alternatives or exclusions (e.g., 'for name search use rpps_search_by_name'). Since the purpose is highly specific, usage is well implied, yet it stops short of explicit routing among siblings, earning a 4 rather than a 5.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server