📋 compliance

Compliance is a gate, not a binder.

8 regulations, each enforced on every pull request by a reviewer agent that reads the rule, your code, and your tests — and blocks the merge if a requirement is missing.

The regulations

One ranking target, one reviewer, every archetype.

📋 PCI-DSS

PCI-DSS compliance

Payment Card Industry Data Security Standard.

📋 HIPAA

HIPAA compliance

Health Insurance Portability and Accountability Act.

📋 GDPR

GDPR compliance

General Data Protection Regulation (EU).

📋 EU AI ACT

EU AI Act compliance

European Union Artificial Intelligence Act.

📋 SOX

SOX compliance

Sarbanes-Oxley Act (Section 404 ITGC).

📋 SOC 2

SOC 2 compliance

Service Organization Control 2 (Trust Services Criteria).

📋 OWASP LLM TOP 10

OWASP LLM Top 10 compliance

OWASP Top 10 for Large Language Model Applications.

📋 21 CFR PART 11

21 CFR Part 11 compliance

FDA Electronic Records & Electronic Signatures.

How it works

Detect the archetype, overlay the reviewer, log the evidence.

GreatCTO detects your project's archetype from its stack and manifests, overlays the matching reviewer agent, and attaches the relevant gates. Every gate decision lands in .great_cto/gates.log — timestamp, reviewer, verdict, rationale — so the audit trail writes itself. Free, MIT, runs locally.

Related deep-dives

More from the blog

AI

What $1.4M of compliance work looks like in 14 hours – ten packs, ten regulated industries

Startups have often reached out to me with the same problem: their team could ship a regulated feature in days, but the compliance setup aro

AI

Three days of code, six weeks of compliance — the math behind why

Not a complaint about lawyers. A breakdown of where the six weeks actually go, and which parts of it are mechanical.

AI

Real cost breakdown: 10 packs, $0.60 LLM bill, $42K saved per regulated feature

Per-feature, per-MVP, per-quarter numbers. Hardware ratios, runway math, and the honest places where the savings stop.

AI

How GreatCTO chooses which compliance pack to attach

Regex vs LLM-based archetype detection, the false-positive count, and why I keep rejecting the obvious fix.