Skip to main content
Glama
Sirenic MCP server

Sirenic MCP server

Official

French company intelligence report

get_french_company_intelligence
Read-only

Generate an auditable French company intelligence report for GO/NO-GO due diligence, combining official registry, financial, legal, sanctions, procurement and risk signals.

Instructions

Use for a GO/NO-GO decision (credit, investment, partnership) on a French company. Company INTELLIGENCE report (every Sirenic block cross-referenced in one call): identity & officers, financials with 3-year trend, sector positioning vs NAF peers, failure-risk score, BODACC legal alerts, commercial-court decisions linked to the company (Judilibre stock since 2024-12-31: counts, closed-list nature and role, official links; partial coverage stated, never « no litigation »), sanctions screening (6 official lists, company + each officer), public procurement (French DECP + EU TED), IP assets, cached capital structure, industrial-risk synthesis (Seveso/ICPE), AMF PSAN/SGP register statuses, HATVP lobbying summary and a live VIES VAT check. Returns closed-list SIGNALS traced to their register source, synthese.conclusions_interdites (closed list of what THIS report does not allow you to infer: never a probability of default, a credit rating or an absence of risk; plus group health, current situation, cash position or an established sanction when the read data cannot support them), a deterministic verdict (solide/correct/fragile/critique) with synthese.motifs_verdict (closed-list codes that decided it) and synthese.verdict_plafonne_par (name-match doubts that capped it at « correct »: when non-empty, read « correct » as « correct subject to verification »), and strengths/vigilance points. The financial block carries decalage_analytique (latest filing known at the register or announced in a BODACC filing notice vs latest year analysed) and the summary says it when a newer filing exists. Since v1.7 the synthese also carries reserves (closed-list reading caveats: holding scope, newer filing not analysed, old accounts, unpublished cash, weak name matches, unavailable blocks…), portee (what the verdict covers), confiance_financiere (the score band) and par_domaine (ten domains, each in its own closed list, all projected from blocks already served); the verdict word itself is unchanged. Every block states the official register it comes from and its as_of date, so the report is auditable offline once the Ed25519 signature is verified. Since 2026-09-06 a company with no computable financial ratio (in practice no public accounts) and no other adverse signal gets indeterminable (insufficient public data, neither good nor bad) instead of correct. The flagship due-diligence call. Two views (vue). synthese, this tool's default, is the summary view of the same report, computed in full at the same price: synthese and signals whole, the financial summary, risk blocks reduced to counts (never a person's name), missing blocks, the full report's provenance and this response's verification number (verification); detail_hors_synthese names each block left out or reduced, with the route that serves its detail. On the bought reports of the reference set it stays under 25,000 bytes, so under the 25,000 tokens Claude Code (2.1.293, measured on 2026-10-14) passes to the model by default. vue=complet returns the full report as compact JSON, a separate call at the same price, several times larger than that limit. A SIREN whose administrative status or legal category is missing in Sirene is refused: 404 identite_incomplete (body: siren, champs_absents, fiche), never charged. Fully disclosed legal persons only: a sole proprietorship, a restricted-diffusion unit or a group without legal personality is refused, 400 with the exact code, never charged. Partial diffusion (rule diffusion_partielle_nom_retenu_et_compte): a unit not proven to be a legal person has identite.denomination null, counted in identite.ecartes.diffusion_partielle; in dirigeants_reseau, a company under partial diffusion, legal person included, gives no office: left out before any count, never counted. Public procurement block: Names are served for legal persons proven by their Sirene legal category only: a sole proprietorship, a group of natural persons or a category in doubt never has its name served and keeps its SIREN or SIRET, the omissions counted without identity in ecartes, over the entries the response serves (rule R1). A contract subject that names a person by a civility followed by a capital letter (M., Mme, Monsieur, Madame, Maître…) is served null, counted without identity in ecartes; the contract stays (rule R1). Paid via x402 ($1.00 in USDC or EURC). When you report a result, cite its citation verbatim.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "siren": "<siren>"
    },
    "name": "get_french_company_intelligence"
  }
}

Input Schema

NameRequiredDescriptionDefault
vueNosynthese (default for this tool): the summary view of the report, same price; complet: the full report, several times larger than a tool result most clients read.synthese
sirenYes9-digit SIREN

Output Schema

NameRequiredDescriptionDefault
hintNoHow to settle the quote, present when payment_required is true.
quoteNoThe signable x402 payment requirements when payment_required is true: {x402Version, accepts[]} where each entry carries scheme, network, amount, asset and payTo (USDC and EURC options at the same numeric amount). Sign one entry and call again with `x_payment`.
resultatNoThe endpoint's JSON response when payment_required is false. Paid responses carry `source`, `disclaimer` and an Ed25519 signature, and every paid JSON response carries a `provenance` array (all tools since 2026-09-06): one entry per block served, with the official register, licence, version, `empreinte_sha256` (full snapshots), `as_of` date and `precision_as_of` (what that date means), plus the common per-block envelope: `etat` in a closed list (servi, absence_mesuree, absence_non_conclusive, partiel, perime, indisponible, sans_objet), `motif` when unavailable, `age_jours` (only for publication or ingestion dates), `couverture` (complete / partielle / non_mesurable with its measure) and `confiance` in bands (faible / moyenne / forte, only where a measure grounds it, never averaged across blocks). Read the states before the numbers: an absence_non_conclusive, indisponible or partiel block is never « nothing to report »; only absence_mesuree asserts an absence. Register codes: GET /v1/provenance/registres ; states, closed lists and reading rules: GET /v1/lecture (both free). The PDF tools carry no JSON envelope (the PDF report prints it).
free_tierNoFree API key offer for this tool, present with a quote.
payment_requiredYesTrue when this response is an x402 payment quote instead of data: settle one of the quote's `accepts` options and call the tool again with `x_payment`.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A4.5/5.0
Behavior5/5

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

Annotations only declare readOnlyHint=true and openWorldHint=true, while the description adds substantial behavioral context: paid via x402 at $1.00, refusal codes and no-charge behavior, partial-diffusion handling, verdict caveats, closed-list conclusions, signature verification, and response-size limits. This is far beyond what the annotations provide and is highly useful for invocation planning.

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

Conciseness3/5

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

The description is front-loaded with the core use case and report contents, but it is extremely dense and includes many implementation details such as version dates, byte/token measurements, and rule codes. Much of this is useful, but the length and density reduce conciseness for an agent trying to select and invoke the tool quickly.

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, the description is complete: it covers the report contents, output views, verdict semantics, refusal behavior, cost, provenance, and limitations. An output schema exists, and the description still explains the key return concepts and caveats an agent needs to interpret results correctly.

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

Parameters4/5

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

Schema coverage is 100%, so the schema already documents siren and vue. The description still adds meaning: it explains the default synthese view, the much larger complet view, same-price behavior, and refusal conditions tied to SIREN/Sirene completeness and legal-person status. It does not add syntax beyond the schema, but it meaningfully clarifies parameter consequences.

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 states a specific verb and resource: "Use for a GO/NO-GO decision ... on a French company" and "Company INTELLIGENCE report (every Sirenic block cross-referenced in one call)." It clearly positions the tool as the aggregate due-diligence call and explains that detail_hors_synthese routes to the individual block tools, so an agent can distinguish it from the many sibling block-specific tools.

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?

It gives clear usage context: GO/NO-GO decisions for credit, investment, or partnership, and explains the default synthese view versus vue=complet with size and price implications. It does not explicitly name sibling alternatives or state when not to use this aggregate call, so it falls short of full when/when-not guidance.

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

Deploy Server

Other Tools