Install
$ agentstack add skill-the-ai-directory-company-agents-and-skills-compliance-assessment ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Compliance Assessment
Before you start
Gather the following from the user. If anything is missing, ask before proceeding:
- Which frameworks apply? — SOC 2, GDPR, HIPAA, ISO 27001, PCI DSS, or others
- What is the scope? — Which systems, services, data types, and teams are in scope
- What is the driver? — Customer requirement, regulatory obligation, funding milestone, or internal initiative
- What is the timeline? — Audit date, certification deadline, or customer commitment
- What exists today? — Current policies, controls, prior audit reports, known gaps
- Who are the stakeholders? — Engineering, legal, security, executive sponsor
Assessment template
1. Define Scope Boundary
Document exactly what is in and out of scope:
In Scope:
- Systems: [List every system, service, and data store]
- Data types: [PII, PHI, financial data, credentials, logs]
- Teams: [Engineering, DevOps, support, HR]
- Environments: [Production, staging, CI/CD pipelines]
- Third parties: [Cloud providers, SaaS tools, subprocessors]
Out of Scope:
- [System/area]: [Reason]
Rules: if a third-party processes in-scope data, they are in scope. If staging has production data copies, staging is in scope. CI/CD pipelines deploying to production are in scope.
2. Map Controls to Framework Requirements
For each framework, map existing controls to requirements per control area:
| Requirement | Framework Ref | Current Control | Status | Gap | |-------------|--------------|-----------------|--------|-----| | Unique user IDs | SOC 2 CC6.1, ISO A.9.2.1 | SSO via Okta | Compliant | None | | MFA for privileged access | SOC 2 CC6.1, ISO A.9.4.2 | MFA for admin roles | Compliant | None | | Access reviews | SOC 2 CC6.2, ISO A.9.2.5 | None | Gap | Quarterly review needed |
Cover at minimum: Access Control, Data Protection (encryption, key management, retention), Logging and Monitoring, Incident Response, Change Management, Business Continuity, Vendor Management, HR Security (training, onboarding/offboarding).
3. Evidence Inventory
For every "Compliant" or "Partial" control, document: the control name, evidence type (policy document, configuration screenshot, logs/reports, process records, or third-party certification), evidence location, last verified date, and refresh cadence.
If evidence does not exist for a control marked "Compliant," it is not compliant. Undocumented controls fail audits.
4. Gap Analysis
For each gap, document using this format:
Gap ID: GAP-001
Control Area: Access Control
Requirement: Quarterly access reviews (SOC 2 CC6.2)
Current State: No regular access reviews performed
Risk Level: High
Remediation: Implement quarterly review with automated user listing
Evidence Needed: Review records, revocation tickets
Estimated Effort: 2 weeks to implement, 4 hours/quarter ongoing
Risk levels: Critical (regulatory violation, active exposure), High (control missing entirely), Medium (partially implemented), Low (exists but evidence incomplete).
5. Remediation Roadmap
Prioritize gaps into phases. Each item must have a single owner, deadline, and evidence deliverable:
- Phase 1 (0-30 days): Critical and High gaps
- Phase 2 (30-90 days): Medium gaps
- Phase 3 (90-180 days): Low gaps and documentation
6. Framework-Specific Considerations
SOC 2: Type II requires a period of observation (3-12 months). Controls implemented one week before the audit window provide insufficient evidence.
GDPR: Requires data processing records (Article 30), DPIAs for high-risk processing (Article 35), and documented lawful basis for each processing activity.
HIPAA: Requires formal risk analysis, Business Associate Agreements with all PHI vendors, and minimum necessary access. Breach notification within 60 days.
ISO 27001: Requires a formal ISMS with Statement of Applicability, risk treatment plan, and management review. Certification requires accredited external audit.
Quality checklist
Before delivering a compliance assessment, verify:
- [ ] Scope boundary is documented with explicit in/out decisions
- [ ] Every control area has been assessed against each applicable framework
- [ ] Each compliant control has evidence with location and verification date
- [ ] Gaps include risk level, remediation steps, and evidence requirements
- [ ] Remediation roadmap is phased with owners, deadlines, and deliverables
- [ ] Framework-specific requirements (observation periods, mandatory documents) are addressed
- [ ] Third-party/vendor controls are assessed, not assumed compliant
Common mistakes
- Treating compliance as point-in-time. Controls must remain effective after the audit. Build ongoing processes, not one-time fixes.
- Assuming vendor compliance covers you. Your cloud provider's SOC 2 does not make you SOC 2 compliant. You own your configuration.
- Documenting controls that do not exist. Auditors verify operation, not just policy documents. A policy without an implemented process is a finding.
- Ignoring evidence requirements. A control without evidence is a gap. Every control must produce verifiable proof of operation.
- Scoping too narrowly. Excluding systems that process in-scope data creates unassessed risk. Auditors will question exclusions.
- No single owner for remediation. "The team will handle it" means no one will. Every gap needs one person accountable by a specific date.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: The-AI-Directory-Company
- Source: The-AI-Directory-Company/agents-and-skills
- License: MIT
- Homepage: https://ai-directory.company
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.