La interfaz está localizada; las descripciones y documentos de las Skills se traducen automáticamente con IA a su idioma. El código, los comandos, rutas y términos técnicos se mantienen en el original, y prevalece el documento del autor.
Explorar Skills
Documentos · SKILL.md

str-report

str-report

This skill should be used when the user asks to draft or structure STR reports, suspicious transaction reports, SAR, suspicious activity reports, draft STR, STR narrative, file suspicious activity, AML STR, goAML, FinCEN SAR, suspicion narrative, or MLRO report. Guides jurisdiction-agnostic STR/SAR drafting—narrative structure (who, what, when, where, why s…

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentación de la Skill

GitHub

name: str-report description: | This skill should be used when the user asks to draft or structure STR reports, suspicious transaction reports, SAR, suspicious activity reports, draft STR, STR narrative, file suspicious activity, AML STR, goAML, FinCEN SAR, suspicion narrative, or MLRO report. Guides jurisdiction-agnostic STR/SAR drafting—narrative structure (who, what, when, where, why suspicious), red flags and typologies, transaction aggregation and chronology, subject identification fields, supporting documentation checklists, quality review before filing, and escalation to MLRO/compliance—not TM rule building (aml-compliance), full LE case management, legal filing duty determination (commercial-counsel), or deep blockchain tracing (blockint skills). Complements aml-compliance, aml-cft, auditor, compliance-engineer, and commercial-counsel.

STR Report

When to Use

  • Draft or restructure a suspicious transaction report (STR) or suspicious activity report (SAR) narrative for AML compliance filing or internal MLRO review
  • Build the who / what / when / where / why suspicion story with clear chronology and aggregated transaction facts
  • Map red flags and typologies (structuring, layering, trade-based ML, etc.) to observable facts—not conclusions without evidence
  • Compile subject and customer identification fields, related parties, accounts, and product/channel context for filing templates
  • Produce a supporting documentation checklist and exhibit index aligned to the narrative
  • Run quality review (completeness, consistency, tone, PII handling) before submission to MLRO or regulator-facing systems
  • Compare jurisdiction-agnostic structure with high-level US (FinCEN SAR), EU, and goAML field concepts (not legal filing advice)
  • Distinguish STR/SAR from internal case notes, management summaries, or law enforcement referral packages
  • Escalate and hand off to MLRO/compliance with a structured fact pack and open questions

When NOT to Use

  • Design transaction monitoring rules, thresholds, alert tuning, or TM scenario libraries → aml-compliance
  • Build a full AML/CFT program, KYC/CDD policy, or enterprise risk assessment → aml-compliance, aml-cft
  • Determine legal obligation to file, statute of limitations, or regulator-specific filing deadlines as counsel → commercial-counsel
  • Conduct internal or IT audit workpapers, SOC 2 testing, or control effectiveness sampling → auditor
  • Implement technical controls, evidence automation, or screening system engineering → compliance-engineer
  • Perform on-chain investigation, wallet tracing, or blockchain intelligence reports → on-chain-investigator-agent, solana-tracing-specialist, blockint skills
  • Draft law enforcement tactical packages, subpoena responses, or victim tracing narratives → route to investigation skills; keep STR factual and filing-oriented
  • CFT/TF-only typologies without broader STR narrative needs → aml-cft (then return here for unified STR assembly)

Related skills

NeedSkill
Full AML program, KYC/CDD, TM scenarios, alert triageaml-compliance
Terrorist financing / proliferation financing STR anglesaml-cft
IT audit workpapers, control testing, samplingauditor
SOC/ISO evidence pipelines, technical control automationcompliance-engineer
Legal interpretation, contracts, filing duty as counselcommercial-counsel
FATF standardized terms (STR, CDD, etc.)fatf-glossary-reference
Transaction screening UI and rescreen conceptstransaction-screening-workflow-concepts
Address screening and list policy conceptsaddress-screening-workflow-concepts
Behavioral monitoring heuristics (educational)behavioral-risk-screening-concepts
Blockchain tracing and forensic reportson-chain-investigator-agent, solana-tracing-specialist

Core Workflows

1. Intake and scope

  1. Confirm jurisdiction, entity, filing type (STR vs SAR vs internal referral), and case ID
  2. Collect alert source, investigation notes, and already-reviewed transactions (do not re-investigate beyond narrative needs)
  3. Identify subjects (customer, beneficial owners, counterparties, unknown parties) and reporting institution role
  4. Flag gaps requiring analyst follow-up before MLRO sign-off

See references/str_report_scope.md.

2. Facts, aggregation, and chronology

  1. Normalize amounts, currencies, dates (UTC vs local—state assumption)
  2. Aggregate related transactions (account, counterparty, corridor, instrument)
  3. Build a chronological timeline with anchors (onboarding, profile change, alert, key txs)
  4. Separate verified facts from inference; label each suspicion element

See references/subject_and_transaction_detail.md.

3. Narrative and typologies

  1. Open with summary suspicion in plain language (one short paragraph)
  2. Develop who / what / when / where / why sections with cross-references to exhibits
  3. Tie red flags to typologies and observable indicators—avoid legal conclusions
  4. Address innocent explanations considered and why rejected or unresolved

See references/narrative_structure_and_quality.md and references/red_flags_and_typologies.md.

4. Format, review, and handoff

  1. Map narrative blocks to jurisdiction template fields at a high level (FinCEN SAR, goAML, etc.)
  2. Complete supporting documentation checklist and quality gates
  3. Run MLRO/compliance escalation pack: decision requested, risks, open items
  4. Record version, reviewers, and not legal advice disclaimer where appropriate

See references/jurisdiction_formats_and_filing.md and references/review_escalation_and_handoff.md.

Output Standards

  • Use past tense for facts, present tense for ongoing risk where helpful; avoid sensational language
  • Every suspicion statement should cite specific transactions, dates, amounts, or documents
  • Minimize PII in drafts shared broadly; full detail in MLRO/regulator packages per policy
  • Include disclaimer: output supports compliance drafting—not legal advice on whether or when to file
  • Deliver: executive summary, full narrative, chronology table, subject/account appendix, exhibit checklist, open questions

Reference Files

FileUse when
references/str_report_scope.mdBoundaries, definitions, STR vs SAR vs internal notes
references/narrative_structure_and_quality.mdWho/what/when/where/why, tone, completeness
references/red_flags_and_typologies.mdIndicators, typology mapping, fact-to-suspicion links
references/subject_and_transaction_detail.mdIDs, accounts, aggregation, chronology tables
references/jurisdiction_formats_and_filing.mdFinCEN SAR, goAML, EU concepts (high level)
references/review_escalation_and_handoff.mdQA gates, MLRO pack, escalation
Skills guardadas