Skip to main content

workflow-kyc-customer-intake

WHAT: End-to-end KYC customer onboarding pipeline that collects identification documents and questionnaire data, coordinates UBO verification, runs `kyc_risk_assessment` (FATF 5-dimension scoring), determines the DD tier (SDD/CDD/EDD), and produces an auditable KYC file ready for MLRO sign-off. WHEN: Invoke when onboarding a new investor (individual, corporate, fund-of-funds, trust, foundation) or a new counterparty (prime broker, lender, vendor, distribution partner) and a complete risk-rated KYC file must be produced.

来源信息

仓库
fall-development-rob/corp_finance
最近来源活动
2026年5月11日 14:10
检测到的 SKILL.md 语言
英语
星标
8
Forks
2

安装方式

默认使用会先检查来源的 Prompt;你也可以切换为直接命令,或下载本地副本。

检查来源文件

决定是否安装前,请先阅读 SKILL.md,以及 SkillsMP 当前展示的配套文件。

正在显示 SKILL.md

SKILL.md
来源说明 · 只读预览
name
workflow-kyc-customer-intake
description
WHAT: End-to-end KYC customer onboarding pipeline that collects identification documents and questionnaire data, coordinates UBO verification, runs `kyc_risk_assessment` (FATF 5-dimension scoring), determines the DD tier (SDD/CDD/EDD), and produces an auditable KYC file ready for MLRO sign-off. WHEN: Invoke when onboarding a new investor (individual, corporate, fund-of-funds, trust, foundation) or a new counterparty (prime broker, lender, vendor, distribution partner) and a complete risk-rated KYC file must be produced.
# KYC — Customer Intake and Onboarding ## What this skill covers The master onboarding pipeline that orchestrates all KYC sub-workflows into a single risk-rated, auditable KYC file. Coordinates identification document collection, UBO verification, sanctions/PEP screening, country risk classification, FATF-aligned risk scoring, DD tier determination, source of funds verification, and ongoing monitoring setup. ## Core principle Documentation is the deliverable. A KYC file that cannot be reconstructed by an auditor is a failed file, even if the underlying conclusion is correct. ## Tool references | Tool | Use | |------|-----| | `kyc_risk_assessment` | FATF 5-dimension risk scoring: customer type (25), geographic (25), product (20), transaction (15), source of wealth (15) | | `sanctions_screening` | Levenshtein fuzzy matching against OFAC/EU/HMT/UN/FATF lists (see `workflow-kyc-sanctions-screening`) | | `country_risk_assessment` | 12-factor sovereign risk + regulatory overlay (see `workflow-kyc-beneficial-ownership` and `workflow-kyc-monitoring-triggers`) | ## Document collection requirements | Customer type | Required documents | |---------------|--------------------| | Individual | Government-issued photo ID, proof of address (<3 months old: utility bill/bank statement), tax ID | | Corporate | Certificate of incorporation, register of directors, register of shareholders, articles, proof of registered address | | Trust | Trust deed, list of trustees/settlors/beneficiaries, registered address | | Fund | PPM, investment management agreement, regulator registration evidence | ## Workflow ### Step 1 — Collect identification documents Gather all required documents per the customer type table above. Record: - Document type, issuing authority, issue/expiry date, document reference. - Any document expired or missing → intake is blocked until resolved. ### Step 2 — Collect KYC questionnaire Capture from the customer: - Customer type (individual / corporate / fund / trust). - Primary jurisdiction of residence/incorporation. - Business activity or occupation. - Expected investment size and frequency. - Source of funds (origin of the specific transaction). - Source of wealth (origin of total accumulated wealth — required for CDD and EDD). - Transaction profile: frequency, channel (wire/custody), geographic counterparties. ### Step 3 — Identify UBOs and run UBO Verification For every entity customer: invoke `workflow-kyc-beneficial-ownership` for each natural person owning ≥25% directly or via aggregated indirect ownership, and for every controller (board majority, contractual veto, special voting share). The highest individual UBO risk rating drives entity-level uplift. ### Step 4 — Sanctions screening Invoke `workflow-kyc-sanctions-screening` for: the customer entity, every UBO, every director, and every authorised signatory. A true positive blocks onboarding regardless of all other factors. ### Step 5 — PEP screening Invoke `workflow-kyc-pep-screening` for every UBO, director, and authorised signatory. A confirmed PEP triggers EDD regardless of risk score. ### Step 6 — Country risk classification For the customer's primary jurisdiction, registered office jurisdiction, and the jurisdiction of every UBO: assess country risk as per the country risk workflow embedded in `workflow-kyc-beneficial-ownership`. ### Step 7 — KYC risk scoring Call `kyc_risk_assessment` with the assembled data. The tool applies FATF 5-dimension scoring: | Dimension | Weight | Inputs | |-----------|--------|--------| | Customer type | 25 | Individual vs corporate vs fund; PEP flag | | Geographic | 25 | Primary jurisdiction FATF status, sanctions regime | | Product | 20 | Investment product, complexity, liquidity | | Transaction | 15 | Volume, frequency, channel, geographic reach | | Source of wealth | 15 | Documented vs opaque; cross-corroborated vs single-source | Score interpretation: <30 = low (SDD eligible); 30-70 = medium (CDD); >70 = high (EDD mandatory). ### Step 8 — DD tier determination Apply the DD tier matrix: | Tier | Trigger | |------|---------| | SDD | Score <30, low-risk jurisdiction, regulated counterparty (e.g., listed bank) | | CDD | Default (30-70), neutral jurisdiction, no PEP, no sanctions concerns | | EDD | Score >70, high-risk jurisdiction, confirmed PEP, sanctions-adjacent, complex ownership structure | The DD tier is the highest tier triggered by any single factor. PEP → EDD regardless of score. High-risk jurisdiction in UBO chain → EDD regardless of score. ### Step 9 — Source of funds / source of wealth verification Invoke `workflow-kyc-source-of-funds` with the determined DD tier. ### Step 10 — Set up ongoing monitoring Invoke `workflow-kyc-monitoring-triggers` to define monitoring rules and the next periodic review date. ### Step 11 — MLRO review and sign-off Draft the KYC file. The MLRO posts the final decision (approve / reject / further review). ## Output format — KYC file ``` KYC FILE --------- Customer name: [name] Customer type: [individual / corporate / fund / trust] Primary jurisdiction: [country] Intake date: [DD-Mon-YYYY] DD tier: [SDD / CDD / EDD] MLRO clearance status: [pending / approved / rejected] 1. Identification documents [list with type, date, reference] 2. KYC questionnaire summary 3. UBO chain [natural persons ≥25% / control; per `workflow-kyc-beneficial-ownership`] 4. Sanctions screening table [per `workflow-kyc-sanctions-screening`] 5. PEP screening table [per `workflow-kyc-pep-screening`] 6. Country risk table [per jurisdiction] 7. KYC risk score [overall + per-dimension] 8. DD tier determination [tier + triggering factors] 9. SoF / SoW evidence list [per `workflow-kyc-source-of-funds`] 10. Ongoing monitoring rules and next review date [per `workflow-kyc-monitoring-triggers`] 11. MLRO sign-off block: "Prepared by: [date] | Reviewed by MLRO: | Decision: approve / reject / further review" ``` ## Quality gates - Every UBO ≥25% identified to natural-person level — no corporate UBO accepted as terminus. - All four screening types run (sanctions + PEP + country risk + KYC risk score) before file is complete. - DD tier cites the highest-triggering factor — not an averaged judgement. - EDD files carry two-source SoF/SoW corroboration. - MLRO sign-off precedes onboarding completion. - File is reconstructable by a regulator from documented evidence alone. ## Related skills - `workflow-kyc-beneficial-ownership` — UBO chain verification (Step 3) - `workflow-kyc-sanctions-screening` — OFAC/EU/HMT/UN/FATF screening (Step 4) - `workflow-kyc-pep-screening` — PEP categories 1-4 (Step 5) - `workflow-kyc-source-of-funds` — SDD/CDD/EDD documentation tier (Step 9) - `workflow-kyc-monitoring-triggers` — ongoing monitoring setup (Step 10) ## Routing **Primary agent:** `cfa-esg-regulatory-analyst`
在 GitHub 查看