Company intelligence report — every block cross-referenced, closed-list signals, rule-based verdict

Identity, officers, financials with 3-year trend, sector positioning (NAF benchmarks), failure-risk score, BODACC alerts, sanctions screening (6 lists, company + officers), procurement (French DECP + EU TED), IP, cached capital structure, industrial-risk synthesis (Seveso/ICPE), AMF PSAN/SGP registers, HATVP lobbying summary, live VIES VAT check — plus closed-list signals traced to their register source, a deterministic verdict (solide/correct/fragile/critique/indeterminable) with `synthese.motifs_verdict` (closed-list codes that decided it) and `synthese.verdict_plafonne_par` (name-match doubts that capped a would-be « solide » at « correct » — when non-empty, read « correct » as « correct subject to verification »; v1.6), `synthese.reserves` / `portee` / `confiance_financiere` / `par_domaine` (v1.7: closed-list reading caveats, scope, score confidence band and a ten-domain projection of blocks already served — the verdict word is unchanged), and strengths/vigilance points. Ed25519-signed (v1.2). Carries the common `provenance[]` envelope (states and reading rules: GET /v1/lecture, free). AMF blacklist (since 2026-09-04): `risques.alertes_amf` serves the matches with score, level and qualification from the shared screening analyser; the vigilance signal `correspondance_liste_noire_amf` is raised only when a match weighs in the concordance table, otherwise the informative `rapprochement_liste_noire_amf_partiel` is emitted and no vigilance point is added. Since 2026-09-06 a failure score in the « vigilance » band raises `score_defaillance_vigilance` — a vigilance point and the verdict motif (negative, not damning: caps at « correct »), so a « correct » decided by the score alone is never served with an empty `points_vigilance`; a « correct » with no vigilance point now reads « solid not granted, see motifs_verdict ». Preserve coverage and provenance. Missing data is not zero; a name match is not proof of identity.

price-ascending: candidates are ordered by the listed price of the operation, with no managed-first preference applied. Unlike /v1/apis, this surface does not rank Apiosk-settled listings above federated ones — a comparison that reorders on commercial grounds is not a comparison. Pass sort=managed_first for the catalog's ordering.

Compared: listed price, settlement rail and input compatibility. Not measured yet: latency, reliability, result quality, and provider terms such as rate limits, jurisdiction and licensing.