transaction-monitoring
>
Works with
--- name: transaction-monitoring description: > license: MIT --- # Transaction Monitoring Skill - v1.0.0 AML transaction monitoring for ongoing customer relationships. Identifies suspicious patterns, assesses structuring and layering, and prepares Suspicious Activity Reports (SARs). ## Overview Transaction monitoring is a continuous obligation under AML regulations. This skill analyzes transaction data to detect patterns that may indicate money laundering, terrorist financing, or other financial crime. ## Detection Patterns ### 1. Structuring (Smurfing) - Multiple transactions just below reporting thresholds - Multiple deposits across different branches/channels on same day - Pattern of round-number transactions - **Threshold (US):** $10,000 CTR threshold; $5,000 aggregated suspicious activity ### 2. Layering - Complex routing through multiple accounts or entities - Rapid movement between accounts with no clear business purpose - International transfers through multiple jurisdictions - Use of shell companies or nominee accounts ### 3. Circular Flows - Money leaving and returning to same account via intermediaries - Round-trip transactions with no economic substance - Funds flowing through subsidiaries back to parent ### 4. Velocity Anomalies - Sudden increase in transaction frequency - Dramatic change in average transaction value - Activity inconsistent with customer profile or declared business - Dormant account reactivation with high-value activity ### 5. Geographic Risk Patterns - Transactions to/from FATF grey-list or black-list jurisdictions - Unexplained payments to/from high-risk countries - Wire transfers to known shell company jurisdictions ## Monitoring Process ### Step 1: Data Analysis - Import transaction data (CSV, Excel, or database extract) - Calculate baseline metrics (average value, frequency, counterparties) - Identify deviations from customer profile ### Step 2: Pattern Detection - Apply detection rules for each pattern type - Flag transactions meeting alert criteria - Score alerts by severity (LOW/MEDIUM/HIGH/CRITICAL) ### Step 3: Investigation [HITL Checkpoint] - Analyst reviews flagged transactions - Assesses whether activity has legitimate explanation - Researches counterparties and destinations - Documents investigation findings ### Step 4: Decision - **CLEAR:** Document rationale, close alert - **SUSPICIOUS:** Draft SAR narrative, escalate to MLRO - **CRITICAL:** Immediate MLRO notification, consider account freeze ### Step 5: SAR Preparation (if applicable) Claude drafts SAR narrative including: - Subject information - Suspicious activity description - Transaction details - Why activity is suspicious - Supporting evidence **CRITICAL HITL: SAR filing is always a human decision. Claude drafts; MLRO reviews and files.** ## Output - Transaction analysis report (Excel with flagged items) - Alert summary with severity scoring - SAR draft (if suspicious activity identified) - Escalation brief for MLRO - Updated customer risk assessment ## Regulatory Basis - **US:** BSA/AML, FinCEN SAR requirements ($5,000 threshold) - **UK/EU:** AMLD5 ongoing monitoring, POCA 2002 SAR obligations - **MENA:** CBUAE, SAMA transaction monitoring requirements --- **Version:** 1.0.0 **Author:** Vyayasan
More Observability skills
google-agents-cli-observability
google/agents-cli
>
azure-observability
microsoft/azure-skills
Azure Observability Services including Azure Monitor, Application Insights, Log Analytics, Alerts, and Workbooks. Provides metrics, APM, distributed tracing, KQL queries, and interactive reports. USE FOR: Azure Monitor, Application Insights, Log Analytics, Alerts, Workbooks, metrics, APM, distributed tracing, KQL queries, interactive reports, observability, monitoring dashboards. DO NOT USE FOR: instrumenting apps with App Insights SDK (use appinsights-instrumentation), querying Kusto/ADX clusters (use azure-kusto), cost analysis (use azure-cost-optimization).
workers-best-practices
cloudflare/skills
Reviews and authors Cloudflare Workers code against production best practices. Load when writing new Workers, reviewing Worker code, configuring wrangler.jsonc, or checking for common Workers anti-patterns (streaming, floating promises, global state, secrets, bindings, observability). Biases towards retrieval from Cloudflare docs over pre-trained knowledge.

