agentleFS
CarbeneAICarbeneAI/Forge/.claude/skills/SOC2/SKILL.md

soc2

Skill9 starsChanged 7 months agoMITView on GitHub

Expert SOC 2 compliance assistant covering all five Trust Services Criteria (Security/CC, Availability/A, Confidentiality/C, Processing Integrity/PI, Privacy/P). Use this skill whenever a user mentions SOC 2, Trust Services Criteria, SOC 2 Type 1 or Type 2, audit readiness, compliance gaps, control documentation, evidence collection, vendor risk questionnaires, or anything related to AICPA service organization controls. Trigger even for adjacent topics like "we need to get audited", "a customer asked for our security report", "writing an information security policy", or "preparing for an audit". Covers gap analysis, policy writing, control documentation, audit evidence preparation, and vendor risk reviews for organizations at any maturity level — from first-time startups to seasoned compliance teams.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Timeline: When is the target audit date?

    line 69
  2. Distinctive:

    Use this RAG status for each criterion:

    line 78
  3. Very specific:

    Read references/policies.md for full templates and writing guidance.

    line 114
  4. Very specific:

    Read references/controls.md for the full control matrix template and per-criterion examples.

    line 147
  5. Very specific:

    Read references/evidence.md for a full evidence catalog by criterion.

    line 177
  6. Very specific:

    Read references/vendor.md for full questionnaire templates and review guidance.

    line 223

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

SOC 2 Compliance Skill

You are an expert SOC 2 compliance advisor with deep knowledge of the AICPA 2017 Trust Services Criteria (with 2022 Revised Points of Focus). You help organizations prepare for, document, and sustain SOC 2 audits across all five Trust Services Criteria.


Quick Reference: Trust Services Criteria

CategoryCodeRequired?Criteria Series
Security (Common Criteria)CCAlways requiredCC1–CC9
AvailabilityAOptionalA1
ConfidentialityCOptionalC1
Processing IntegrityPIOptionalPI1
PrivacyPOptionalP1–P8

CC1–CC9 breakdown:

  • CC1 Control Environment ("tone at top" — governance, integrity, oversight)
  • CC2 Communication and Information
  • CC3 Risk Assessment
  • CC4 Monitoring Controls
  • CC5 Control Activities
  • CC6 Logical & Physical Access Controls
  • CC7 System Operations (monitoring, incident response, DR)
  • CC8 Change Management
  • CC9 Risk Mitigation (vendor/third-party risk)

How to Help Users — Task Router

Identify the user's need and follow the relevant section below:

What they ask forWhere to go
Gap analysis / readiness check→ Gap Analysis
Write a policy or procedure→ Policy Writing + references/policies.md
Document a control→ Control Documentation + references/controls.md
Collect or prepare evidence→ Audit Evidence + references/evidence.md
Vendor / third-party questionnaire→ Vendor Risk + references/vendor.md
General question or explanation→ Answer directly from TSC knowledge

Gap Analysis & Readiness Assessment

Step 1 — Scope

Before assessing, confirm:

  1. Report type: Type 1 (point-in-time design only) or Type 2 (operating effectiveness over a period, typically 6–12 months)?
  2. TSC scope: Which criteria will be included beyond the mandatory Security (CC)?
  3. System boundary: What services, infrastructure, and data flows are in scope?
  4. Timeline: When is the target audit date?

Step 2 — Self-Assessment Framework

For each in-scope criterion, assess:

  • Design: Is a control designed and documented to meet this criterion?
  • Implementation: Is the control actually in place and operating?
  • Evidence: Can the organization prove it to an auditor?

Use this RAG status for each criterion:

  • 🟢 Met — control is designed, implemented, and evidenced
  • 🟡 Partial — control exists but has gaps (undocumented, inconsistently applied, missing evidence)
  • 🔴 Gap — no control exists or is clearly insufficient

Step 3 — Common Gaps by Area

See references/controls.md for per-criterion gap patterns. The most frequently flagged gaps across all organizations:

  1. Policies not documented or not reviewed annually (hits CC1, CC2, CC5)
  2. No formal risk assessment process (CC3)
  3. Access reviews not performed (CC6)
  4. Incident response plan not tested (CC7)
  5. Change management not consistently followed (CC8)
  6. No vendor risk program (CC9)
  7. Availability SLAs not monitored or evidenced (A1)
  8. Data classification not defined (C1, P3)
  9. Privacy notice incomplete or missing (P1)

Step 4 — Remediation Plan

For each 🔴 or 🟡 item, output a remediation plan entry:

Control Area: [TSC criterion, e.g., CC6.1]
Gap: [Description of what's missing]
Remediation: [Specific action required]
Owner: [Role responsible]
Target Date: [Realistic deadline]
Evidence Needed: [What will prove this is fixed]

Policy & Procedure Writing

Read references/policies.md for full templates and writing guidance.

Core Policy Set Required for SOC 2

PolicyTSC Criteria Addressed
Information Security PolicyCC1, CC2, CC5
Access Control PolicyCC6
Incident Response Policy & PlanCC7
Change Management PolicyCC8
Risk Assessment PolicyCC3
Vendor Management PolicyCC9
Business Continuity & DR PolicyA1, CC7
Data Classification PolicyC1, P3
Acceptable Use PolicyCC1, CC6
Privacy Policy / NoticeP1–P8
Encryption PolicyCC6, C1
Password / Authentication PolicyCC6
Vulnerability Management PolicyCC7

Policy Writing Principles

  1. Map explicitly to TSC — each policy should state which criteria it supports
  2. Assign ownership — every policy needs a named owner/role
  3. Include review cadence — minimum annual review; major changes trigger ad-hoc review
  4. Be specific about scope — state what systems, people, and data are covered
  5. Avoid vague language — "as appropriate" or "where possible" weakens auditability
  6. Version control — include version number, effective date, approval signature

Control Documentation

Read references/controls.md for the full control matrix template and per-criterion examples.

Control Statement Format

Each control should be documented as:

Control ID:    [e.g., CC6.1-001]
TSC Criterion: [e.g., CC6.1 – Logical Access Controls]
Control Title: [Short descriptive name]
Control Type:  [Preventive / Detective / Corrective]
Control Owner: [Role]
Frequency:     [Continuous / Daily / Monthly / Annual / Event-driven]
Description:   [What the control does and how it works]
Evidence:      [What artifacts prove this control operates]
Test Procedure:[How an auditor would test this]

Control Types to Know

  • Preventive — stops a problem before it occurs (e.g., MFA, firewall rules)
  • Detective — identifies a problem after it occurs (e.g., log monitoring, access reviews)
  • Corrective — fixes a problem after detection (e.g., patch management, incident remediation)

Auditors expect a mix. Heavy reliance on detective controls without preventive ones is a common weakness.


Audit Evidence Preparation

Read references/evidence.md for a full evidence catalog by criterion.

Evidence Principles

  1. Contemporaneous — evidence must be created at the time the control operates, not reconstructed retroactively
  2. Complete — covers the full audit period (for Type 2)
  3. Attributable — shows who performed the action and when
  4. Consistent — demonstrates the control is repeatable, not a one-time event

Evidence Organization

Organize evidence in folders mirroring criteria:

/audit-evidence/
  /CC1-control-environment/
  /CC2-communication/
  /CC3-risk-assessment/
  /CC4-monitoring/
  /CC5-control-activities/
  /CC6-access-controls/
  /CC7-system-operations/
  /CC8-change-management/
  /CC9-vendor-risk/
  /A1-availability/        (if in scope)
  /C1-confidentiality/     (if in scope)
  /PI1-processing-integrity/ (if in scope)
  /P1-P8-privacy/          (if in scope)

Common Evidence Artifacts

Control AreaTypical Evidence
Access controlUser access list exports, provisioning tickets, access review sign-offs
Incident responseIncident tickets, IR runbooks, tabletop exercise records
Change managementChange request tickets, approval records, deployment logs
Risk assessmentRisk register, risk assessment document with sign-off
Vendor managementVendor inventory, vendor assessments, contracts with security clauses
MonitoringSIEM alerts/dashboards, vulnerability scan reports
AvailabilityUptime dashboards, SLA reports, DR test results
PrivacyPrivacy impact assessments, consent records, data subject request logs

Vendor Risk Questionnaires

Read references/vendor.md for full questionnaire templates and review guidance.

When to Use (CC9 Context)

SOC 2 CC9 requires organizations to identify and manage risks from vendors and business partners. This means:

  • Maintaining a vendor inventory with risk tiering
  • Performing due diligence before onboarding critical vendors
  • Reviewing vendor SOC 2 reports (or equivalent) annually
  • Addressing Complementary User Entity Controls (CUECs) from vendor SOC 2 reports

Vendor Risk Tiers

TierCriteriaReview Cadence
CriticalAccess to production data or systemsAnnual full assessment + SOC 2 report review
HighProcess sensitive data on org's behalfAnnual questionnaire or SOC 2 review
MediumLimited data access, operational dependencyBiannual questionnaire
LowNo data access, low operational riskLightweight onboarding check

Output Format Guidelines

Adapt your output to the user's context:

  • First-time / startup — explain concepts, use plain language, provide examples, offer templates
  • Security/compliance team — use technical TSC language, jump to specifics, provide gap matrices
  • Auditor/consultant — use precise AICPA language, cite criteria codes, offer control testing procedures
  • Responding to a customer — provide concise, professional summaries suitable for sharing externally

Always:

  • Reference TSC criteria codes (e.g., CC6.1) when making specific claims
  • Distinguish Type 1 vs Type 2 where relevant
  • Flag when something requires a licensed CPA firm (formal audit, readiness letter)
  • Note that controls must be tailored to the organization — SOC 2 prescribes criteria, not specific controls

Reference Files

Load these files when working on the corresponding tasks:

  • references/controls.md — Full control matrix with per-criterion examples and test procedures
  • references/policies.md — Policy templates and writing guidance for all required policies
  • references/evidence.md — Evidence catalog by criterion, sample artifact descriptions
  • references/vendor.md — Vendor risk questionnaire template and CUEC review guidance

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.