> Source: https://www.usehabeo.com/compliance/soc2-readiness
> Published: 2026-02-01 · Updated: 2026-05-15

Phase 10 control inventory mapped against the Trust Services Criteria.

# SOC 2 Readiness Assessment — Habeo

**Date:** 2026-05-15
**Phase:** Phase 10 (kick-off)
**Target audit:** SOC 2 Type 1, Q1 2027
**Auditor:** TBD (RFP out to three firms)

This document inventories the **Trust Services Criteria** (TSC) controls Habeo has in place as of Phase 10, mapped against the SOC 2 framework. It is the **readiness gap analysis** that precedes engaging an auditor — it tells us what to do next, not what's audited.

---

## TSC Categories Covered

- **CC (Common Criteria)** — required for all SOC 2 reports.
- **A (Availability)** — required for the SaaS context.
- **C (Confidentiality)** — required for the multi-tenant context.
- **P (Privacy)** — not pursued in V1; revisit when handling PII beyond enterprise contact info.
- **PI (Processing Integrity)** — not pursued in V1; revisit when adding transactional financial workflows.

---

## Control Inventory

### CC1 — Control Environment

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC1.1 | Code of conduct | Draft | Internal CoC drafted; needs board adoption when board exists |
| CC1.2 | Security org structure | Solo founder | Functional separation lands when team expands |
| CC1.3 | Onboarding / offboarding workflow | Not started | Q3 2026 (after first hire) |
| CC1.4 | Vendor management | Done | DPAs signed with Vercel, Neon, Clerk, Resend, Axiom, PostHog |
| CC1.5 | Code of ethics enforcement | Implicit | Formalize before audit |

### CC2 — Communication & Information

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC2.1 | Security policy publication | In progress | `/compliance` route added in Phase 10 |
| CC2.2 | Customer comms — incidents | Done | Notification SLA in HECVAT response (§15) |
| CC2.3 | Internal security training | Not started | Q4 2026 |

### CC3 — Risk Assessment

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC3.1 | Risk register | Not started | Q3 2026 — pre-audit |
| CC3.2 | Threat modeling | Informal | Per-phase architectural decisions in DECISIONS.md |
| CC3.3 | Fraud risk consideration | Implicit | RBAC + audit trail address core scenarios |

### CC4 — Monitoring Activities

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC4.1 | Continuous monitoring | Done | Axiom log aggregation; alerts on error spikes |
| CC4.2 | Internal audits | Not started | Q1 2027 — quarterly cadence pre-audit |
| CC4.3 | Vendor performance review | Ad-hoc | Formalize quarterly |

### CC5 — Control Activities

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC5.1 | Segregation of duties (DB roles) | Done | `habeo_owner` (migrations) vs `habeo_app` (runtime) since Phase 1 |
| CC5.2 | Code review | Done | All PRs require approval; main branch protected |
| CC5.3 | Change management | Done | Conventional commits + Phase planning docs + Phase completion reports |
| CC5.4 | Backup procedures | Done | Neon point-in-time recovery, 30-day window, RPO 1h |

### CC6 — Logical & Physical Access

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC6.1 | Logical access — RBAC | Done | Three roles; per-route `requireRole()` + per-row RLS; cross-tenant probe in CI |
| CC6.2 | New user provisioning | Done | Clerk-mediated; self-serve gated by `ALLOW_SELF_SERVE_SIGNUP=true` (off in prod) |
| CC6.3 | User deprovisioning | Done | Soft-delete via `deleted_at`; immediate session invalidation in Clerk |
| CC6.4 | Physical access (data centers) | Inherited | Neon + Vercel SOC 2 reports cover physical |
| CC6.5 | Credentials management | Done | Vercel encrypted env vars; no secrets in repo (.gitignore enforced) |
| CC6.6 | MFA | Done | Required for admin via Clerk |
| CC6.7 | Privileged access | Done | `habeo_owner` connection-string is migration-only; runtime never sees it |

### CC7 — System Operations

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC7.1 | Detection (monitoring + alerting) | Done | Vercel function timeouts logged; PostHog event volumes anomaly-alert |
| CC7.2 | Incident response runbook | Draft | Formal runbook needed pre-audit |
| CC7.3 | Vulnerability identification | Done | Renovate + Snyk; CVE response SLA 24h |
| CC7.4 | Patch management | Done | Dependency PRs auto-opened and merged after CI |
| CC7.5 | Backups | Done | Same as CC5.4 |

### CC8 — Change Management

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC8.1 | Code change approval | Done | Branch protection on `main`; PR review required |
| CC8.2 | DB migration approval | Done | Migrations live in versioned Drizzle files; require PR review |
| CC8.3 | Production deploy gating | Done | CI must pass (typecheck + lint + tests + cross-tenant probe) before Vercel auto-deploy |

### CC9 — Risk Mitigation

| ID | Control | Status | Evidence |
|---|---|---|---|
| CC9.1 | Disaster recovery plan | Draft | DR runbook + quarterly tests |
| CC9.2 | Vendor risk | Done | DPAs in place; subprocessor list in HECVAT §16 |

### A — Availability

| ID | Control | Status | Evidence |
|---|---|---|---|
| A1.1 | Uptime monitoring | Done | Vercel's built-in + a Better Stack ping every 60s |
| A1.2 | Capacity planning | Done | Fluid Compute auto-scales; Neon's autoscaling on |
| A1.3 | Recovery testing | Ad-hoc | Quarterly rehearsal lands pre-audit |

### C — Confidentiality

| ID | Control | Status | Evidence |
|---|---|---|---|
| C1.1 | Data classification | Implicit | Stripped-for-role helper marks financial fields; full data-classification policy is pre-audit work |
| C1.2 | Data retention + disposal | Done | Soft-delete + scheduled hard-delete after 30 days (audit trail retained) |
| C1.3 | Encryption at rest | Done | AES-256 via Neon |
| C1.4 | Encryption in transit | Done | TLS 1.3 only |

---

## Gap Summary

**Done (15 of 41):** Controls implemented + evidence accumulating in the normal course of operating the product.

**Partial (10 of 41):** Implemented informally; needs formalization before audit (policy documents, training materials, recurring tasks).

**Not started (4 of 41):** Most are people-process controls that need a team larger than one before they make sense.

## Pre-Audit Work Estimate

- 6–8 weeks of formalization work (policies, runbooks, training materials)
- 2–3 weeks of evidence collection (audit logs export, change tickets, vendor reviews)
- Auditor engagement: 4–6 weeks observation window for Type 1; 6 months for Type 2

Target: SOC 2 Type 1 report in hand by **Q1 2027**, with Type 2 observation window starting immediately after.

---

This document is updated each phase. Next review: Phase 12 (when the team is sized to support the people-process controls).
