Croise FINESS DREES ↔ SIRENE INSEE V3.11 et calcule un score de cohérence (Sørensen-Dice sur bigrammes) pour chaque SIRET candidat. Utile pour confirmer/infirmer un appariement num_finess ↔ SIRET avant prospection ou cross-check qualité. Logique : 1. Récupère FINESS (raison sociale + adresse libe...
Part of the France Data Mcp server.
Free to start. No card required.
AI agents call reconcilier_finess_sirene to retrieve information from France Data Mcp without modifying any data. This is common in research, monitoring, and reporting workflows where the agent needs context before taking action. Because read operations don't change state, they are generally safe to allow without restrictions -- but you may still want rate limits to control API costs.
Even though reconcilier_finess_sirene only reads data, uncontrolled read access can leak sensitive information or rack up API costs. An agent caught in a retry loop could make thousands of calls per minute. A rate limit gives you a safety net without blocking legitimate use.
Read-only tools are safe to allow by default. No rate limit needed unless you want to control costs.
{
"version": "1",
"default": "deny",
"tools": {
"reconcilier_finess_sirene": {}
}
}See the full France Data Mcp policy for all 36 tools.
These attack patterns abuse exactly the kind of access reconcilier_finess_sirene gives an agent. Each links to the full case and the policy that stops it:
Other read tools across the catalogue. The same approach applies to each: allow, with a rate cap to control cost.
Croise FINESS DREES ↔ SIRENE INSEE V3.11 et calcule un score de cohérence (Sørensen-Dice sur bigrammes) pour chaque SIRET candidat. Utile pour confirmer/infirmer un appariement num_finess ↔ SIRET avant prospection ou cross-check qualité. Logique : 1. Récupère FINESS (raison sociale + adresse libellée) 2. Récupère SIRET candidats via la table RPPS 3. Pour chaque SIRET, lookup SIRENE puis calcule 3 sous-scores : - nom : Dice sur raison sociale (FINESS vs SIRENE.uniteLegale) - adresse : Dice sur adresse complète - telephone : binaire 0/1 (toujours 0 actuellement : SIRENE n'expose pas le tel) 4. Score global = pondération (nom 0.5, adresse 0.4, tel 0.1) 5. Verdict brut : match (≥0.8) / partial (0.5..0.8) / mismatch (<0.5) Algorithme PUBLIC (Sørensen-Dice est dans la littérature depuis 1948). Aucune valeur ajoutée Unilabs ici — c'est une primitive ouverte. La connaissance propriétaire (mapping enseignes ↔ SELAS) reste côté Geo Intel. Format : objet LookupResult. Quand found: true, retourne { num_finess, candidates, skipped } : - candidates : tableau trié par score_global décroissant (meilleur match en premier) - skipped : SIRET candidats qu'on n'a PAS pu réconcilier (lookup SIRENE rejected ou not_found) avec la reason. Permet au caller de distinguer 'aucun SIRET candidat trouvé' (found: false LookupResult.not_found) de 'N SIRETs candidats mais tous rejetés par SIRENE' (candidates: [] + skipped: [...]).. It is categorised as a Read tool in the France Data Mcp MCP Server, which means it retrieves data without modifying state.
Register the France Data MCP server in PolicyLayer and add a rule for reconcilier_finess_sirene: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches France Data Mcp. Nothing to install.
reconcilier_finess_sirene is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the reconcilier_finess_sirene rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for reconcilier_finess_sirene. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
reconcilier_finess_sirene is provided by the France Data MCP server (france-data-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 36 France Data Mcp tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4,600+ MCP servers and 31,000+ tools scanned and risk-classified.