Install any skill in seconds. Free to start, no credit card required.
Get Started Free →SOC 2 Type I and Type II compliance management against the Trust Services Criteria. Use for SOC 2 readiness assessments, TSC gap analysis, audit evidence collection, infrastructure control validation, and CPA firm audit preparation.
.claude/skills/borghei-soc2-compliance-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 126% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 89% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 281% | 0% |
SOC 2 Type I and Type II compliance management covering all Trust Services Criteria (TSC), infrastructure security validation, evidence collection, and end-to-end audit preparation.
| Aspect | Type I | Type II | |--------|--------|---------| | Scope | Design of controls at a point in time | Design AND operating effectiveness over a period | | Duration | Single date (snapshot) | Observation period (3-12 months, typically 6-12) | | Cost | $20K-$60K (first audit) | $40K-$150K (first audit) | | Timeline | 1-3 months | 6-15 months (includes observation period) | | Customer Preference | Early-stage acceptable | Enterprise customers require |
Start with Type I to validate control design, then transition to Type II within 6 months.
| Category | Focus | Controls | |----------|-------|----------| | CC1-CC5 | Common Criteria (COSO-based) | Control environment, communication, risk, monitoring, control activities | | CC6 | Logical and Physical Access | Authentication, authorization, physical security, encryption | | CC7 | System Operations | Vulnerability management, monitoring, incident response, BCP | | CC8 | Change Management | Authorization, testing, deployment controls | | CC9 | Risk Mitigation | Vendor management, business disruption, risk transfer | | A1 | Availability | Capacity planning, DR, recovery testing | | PI1 | Processing Integrity | Data validation, error handling, reconciliation | | C1 | Confidentiality | Classification, encryption, disposal | | P1 | Privacy | Notice, consent, data subject rights, retention |
For detailed control requirements per category, see REFERENCE.md.
Before running the readiness assessment, confirm these inputs. If any is unknown or vague, ASK — do not assume:
Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the assessment.
The agent guides organizations through SOC 2 readiness from gap analysis through audit completion.
scripts/soc2_readiness_checker.py.scripts/soc2_infrastructure_auditor.py; evidence collection producing artifacts.| TSC | Evidence Type | Collection Method | Refresh | |-----|---------------|-------------------|---------| | CC1 | Code of conduct acknowledgments | HR system export | Annual | | CC2 | Security awareness training records | LMS export | Ongoing | | CC3 | Risk assessment report, risk register | GRC platform | Annual/Quarterly | | CC4 | Penetration test reports, vulnerability scans | Third-party/scanner | Annual/Monthly | | CC5 | Policy documents with version history | Policy management | Annual review | | CC6 | Access reviews, MFA enrollment, offboarding | IAM/IdP/HRIS | Quarterly/Per event | | CC7 | Vulnerability remediation, incident records | Ticketing/ITSM | Ongoing | | CC8 | Change tickets with approvals, code reviews | ITSM/Git | Per change | | CC9 | Vendor risk assessments, vendor SOC 2 reports | GRC platform | Annual | | A1 | Uptime reports, DR tests, backup logs | Monitoring/backup | Monthly/Semi-annual | | PI1 | Data validation/reconciliation reports | Application logs | Per process | | C1 | Data classification inventory, encryption configs | Manual/automated | Annual/Quarterly | | P1 | PIAs, DSR response tracking | Privacy tool | Per event |
bash# Generate evidence checklist for all TSC categories python scripts/evidence_collector.py --generate-checklist --categories all # Track evidence status python scripts/evidence_collector.py --status evidence-tracker.json # Update specific evidence item python scripts/evidence_collector.py --update evidence-tracker.json \ --item CC6.1-MFA --status collected # Generate readiness dashboard python scripts/evidence_collector.py --dashboard evidence-tracker.json # Export for auditor review python scripts/evidence_collector.py --export evidence-tracker.json --format json
GRC Platforms: Vanta, Drata, Secureframe, Laika, AuditBoard -- automated evidence collection via API integrations, continuous control monitoring, auditor collaboration portals.
Infrastructure-as-Evidence: Cloud configuration snapshots (AWS Config, Azure Policy, GCP Org Policies), Terraform state as configuration evidence, Git history as change management evidence, CI/CD pipeline logs as deployment control evidence.
The agent validates infrastructure configurations against SOC 2 requirements.
| Domain | Key Checks | SOC 2 Mapping | |--------|-----------|---------------| | Cloud (AWS/Azure/GCP) | Encryption, IAM, logging, network, backup, secrets | CC6, CC7, A1, C1 | | DNS | SPF, DKIM, DMARC, DNSSEC, CAA | CC6.6, CC2.2 | | TLS/SSL | TLS 1.2+, AEAD ciphers, HSTS, auto-renewal | CC6.7 | | Endpoint | MDM, disk encryption, EDR, patching, screen lock | CC6.1, CC6.8, CC7.1 | | Network | Segmentation, WAF, DDoS, VPN/ZTNA, egress filtering | CC6.6, A1.1 | | Container | Image scanning, minimal base, no privileged, RBAC | CC6.1, CC7.1 | | CI/CD | Signed commits, branch protection, SAST/DAST, SBOM | CC7.1, CC8.1 | | Secrets | Vault storage, rotation policies, git scanning | CC6.1 |
For detailed per-provider control mappings, see REFERENCE.md.
bash# Full infrastructure audit python scripts/soc2_infrastructure_auditor.py --config infra-config.json # Audit specific domains only python scripts/soc2_infrastructure_auditor.py --config infra-config.json \ --domains dns tls cloud # JSON output with severity ratings python scripts/soc2_infrastructure_auditor.py --config infra-config.json --format json # Generate sample configuration template python scripts/soc2_infrastructure_auditor.py --generate-template
| Phase | Duration | Activities | |-------|----------|------------| | Scoping | 2-4 weeks | Define TSC, system boundaries, auditor selection | | Gap Analysis | 2-4 weeks | Assess current controls, identify gaps | | Remediation | 8-16 weeks | Implement missing controls, policies, procedures | | Type I Audit | 2-4 weeks | Point-in-time control design assessment | | Type II Observation | 3-12 months | Controls operate, evidence collected continuously | | Type II Fieldwork | 2-4 weeks | Auditor testing, evidence review, interviews | | Report Issuance | 2-4 weeks | Draft review, management response, final report |
For severity level definitions and breach notification timelines, see REFERENCE.md.
bash# Full readiness assessment python scripts/soc2_readiness_checker.py --config org-controls.json # JSON output for programmatic use python scripts/soc2_readiness_checker.py --config org-controls.json --format json # Check specific TSC categories python scripts/soc2_readiness_checker.py --config org-controls.json \ --categories security availability # Include cloud provider control mapping python scripts/soc2_readiness_checker.py --config org-controls.json --cloud-mapping
bash# Generate checklist and track status python scripts/evidence_collector.py --generate-checklist --categories all python scripts/evidence_collector.py --status evidence-tracker.json python scripts/evidence_collector.py --dashboard evidence-tracker.json
bash# Validate infrastructure against SOC 2 requirements python scripts/soc2_infrastructure_auditor.py --config infra-config.json python scripts/soc2_infrastructure_auditor.py --config infra-config.json --format json
| Document | Description | |----------|-------------| | REFERENCE.md | Detailed TSC controls, infrastructure checks, access control specs, vendor management, training, IRP, BC/DR | | Trust Services Criteria Guide | Complete TSC reference with control objectives and audit questions | | Infrastructure Security Controls | Cloud, DNS, TLS, endpoint, container, CI/CD security configurations | | Audit Preparation Playbook | End-to-end audit prep guide with timelines, checklists, cost estimation |
| Problem | Likely Cause | Resolution | |---------|-------------|------------| | Readiness checker scores are 0% across all categories | Controls JSON missing config_key values or all set to false | Verify the input JSON maps each TSC control to a boolean value under the correct config_key. Run --generate-sample > sample-config.json to see the expected structure. | | Infrastructure auditor reports all checks as "fail" | Infrastructure config JSON is empty or uses wrong key names | Run --generate-template to produce a valid template. Populate DNS, TLS, cloud, endpoint, and other sections with actual infrastructure state. | | Evidence collector checklist missing categories | --categories flag filtering output | Use --categories all to generate the complete checklist. Available categories: security, availability, processing_integrity, confidentiality, privacy. | | Evidence tracker status not updating | Tracker file path incorrect or file not writable | Verify the path passed to --status or --update points to an existing tracker JSON file. Check file permissions. | | Cloud mapping not appearing in readiness report | --cloud-mapping flag not included | Add --cloud-mapping to the readiness checker command to include AWS/Azure/GCP control mappings in the output. | | Type II observation period too short for auditor | Observation period is less than 3 months | Most CPA firms require a minimum 3-month observation period for Type II. A 6-12 month period carries more weight. Plan the observation window during the scoping phase. | | Auditor requests evidence not in the tracker | Evidence catalog does not cover all TSC subcriteria for the selected scope | Supplement the auto-generated checklist with auditor-specific evidence requests. Each CPA firm may have additional requirements beyond the standard TSC evidence items. |
In Scope:
Out of Scope:
| Skill | Integration | |-------|------------| | infrastructure-compliance-auditor | Provides Vanta-level infrastructure checks across cloud, DNS, TLS, endpoints, access controls, and CI/CD that map directly to SOC 2 TSC requirements | | nist-csf-specialist | NIST CSF functions map to SOC 2 TSC categories; use the control mapper to build unified control matrices for organizations pursuing both | | information-security-manager-iso27001 | ISO 27001 Annex A controls provide a management system backbone that satisfies many SOC 2 requirements; shared evidence reduces audit burden | | pci-dss-specialist | PCI DSS requirements overlap with SOC 2 CC6 (access), CC7 (operations), CC8 (change management); shared controls for payment-processing organizations | | gdpr-dsgvo-expert | GDPR requirements align with SOC 2 Privacy (P1) criteria; organizations processing EU personal data can leverage shared privacy controls | | nis2-directive-specialist | NIS2 minimum security measures overlap with SOC 2 security criteria; EU entities can map shared incident response, access control, and encryption controls |
Evaluates organizational controls against SOC 2 Trust Services Criteria with per-category scoring.
| Flag | Required | Description | |------|----------|-------------| | --config | Yes (or --generate-sample) | Path to organization controls JSON file with boolean values for each TSC control | | --format | No | Output format: json for structured output, omit for human-readable text | | --categories | No | Space-separated TSC categories to assess (e.g., security availability). Omit for all. | | --cloud-mapping | No | Include cloud provider (AWS/Azure/GCP) control mappings in the output | | --generate-sample | No | Generate a sample controls JSON template (pipe to file with > sample-config.json) |
Generates evidence collection checklists and tracks evidence gathering status.
| Flag | Required | Description | |------|----------|-------------| | --generate-checklist | No | Generate an evidence collection checklist for the specified categories | | --categories | No | Space-separated TSC categories: security, availability, processing_integrity, confidentiality, privacy, or all | | --status | No | Path to evidence tracker JSON file to display collection status | | --update | No | Path to evidence tracker JSON file to update (use with --item and --status) | | --item | No | Evidence item identifier to update (e.g., CC6.1-MFA) | | --dashboard | No | Path to evidence tracker JSON file to generate a readiness dashboard | | --export | No | Path to evidence tracker JSON file to export | | --format | No | Export format: json for structured output |
Audits infrastructure configurations against SOC 2 requirements with severity-rated findings.
| Flag | Required | Description | |------|----------|-------------| | --config | Yes (or --generate-template) | Path to infrastructure configuration JSON file with DNS, TLS, cloud, endpoint, and other domain settings | | --format | No | Output format: json for structured findings with severity ratings, omit for human-readable text | | --domains | No | Space-separated infrastructure domains to audit (e.g., dns tls cloud). Omit for all domains. | | --generate-template | No | Generate a sample infrastructure configuration template (pipe to file with > infra-config.json) |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 25,878 | 26,243 | +1% | 1 | 1 | 0% | 4,202 | 9,510 | +126% | 0 | 0 | — |
case-02 | fail→pass | 27,635 | 15,762 | -43% | 1 | 1 | 0% | 5,578 | 7,816 | +40% | 0 | 0 | — |
case-03 | fail→pass | 27,858 | 24,404 | -12% | 1 | 1 | 0% | 4,590 | 8,676 | +89% | 0 | 0 | — |
case-04 | fail→pass | 18,482 | 3,371 | -82% | 1 | 1 | 0% | 2,968 | 5,195 | +75% | 0 | 0 | — |
case-05 | fail→pass | 8,434 | 3,119 | -63% | 1 | 1 | 0% | 1,381 | 5,255 | +281% | 0 | 0 | — |
case-06 | fail→pass | 7,957 | 2,582 | -68% | 1 | 1 | 0% | 1,525 | 5,207 | +241% | 0 | 0 | — |
case-07 | fail→pass | 11,370 | 4,579 | -60% | 1 | 1 | 0% | 1,840 | 5,328 | +190% | 0 | 0 | — |
case-08 | pass→pass | 10,196 | 2,154 | -79% | 1 | 1 | 0% | 1,685 | 5,062 | +200% | 0 | 0 | — |
case-09 | fail→pass | 11,326 | 3,087 | -73% | 1 | 1 | 0% | 1,740 | 5,234 | +201% | 0 | 0 | — |
case-10 | fail→pass | 7,605 | 2,975 | -61% | 1 | 1 | 0% | 1,237 | 5,202 | +321% | 0 | 0 | — |
case-11 | pass→pass | 11,215 | 6,167 | -45% | 1 | 1 | 0% | 1,736 | 5,634 | +225% | 0 | 0 | — |
case-12 | pass→pass | 17,303 | 9,222 | -47% | 1 | 1 | 0% | 2,776 | 6,126 | +121% | 0 | 0 | — |
case-13 | pass→pass | 14,526 | 11,801 | -19% | 1 | 1 | 0% | 2,365 | 6,367 | +169% | 0 | 0 | — |
case-14 | fail→pass | 8,730 | 5,289 | -39% | 1 | 1 | 0% | 1,302 | 5,521 | +324% | 0 | 0 | — |
case-15 | pass→pass | 15,021 | 9,524 | -37% | 1 | 1 | 0% | 2,294 | 6,221 | +171% | 0 | 0 | — |
case-16 | fail→pass | 14,319 | 6,074 | -58% | 1 | 1 | 0% | 2,430 | 5,749 | +137% | 0 | 0 | — |
case-17 | pass→pass | 12,449 | 13,924 | +12% | 1 | 1 | 0% | 1,953 | 6,823 | +249% | 0 | 0 | — |
case-18 | pass→pass | 23,062 | 8,844 | -62% | 1 | 1 | 0% | 2,456 | 6,289 | +156% | 0 | 0 | — |
case-19 | fail→pass | 16,127 | 10,373 | -36% | 1 | 1 | 0% | 2,513 | 6,572 | +162% | 0 | 0 | — |
case-20 | fail→fail | 20,212 | 19,140 | -5% | 1 | 1 | 0% | 3,068 | 7,490 | +144% | 0 | 0 | — |
case-21 | pass→pass | 11,755 | 9,036 | -23% | 1 | 1 | 0% | 1,716 | 6,127 | +257% | 0 | 0 | — |
case-22 | pass→pass | 5,538 | 9,734 | +76% | 1 | 1 | 0% | 415 | 5,708 | +1275% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +55 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.