# Cybersecurity AI Workspace — Threat Lookup, AppSec & SOC | Utah

> Cybersecurity AI built for security engineers: CVE & CWE lookup, OWASP code review, threat hunting, SOC triage, Attack/Defense agents, CLI, 15-tool MCP, API.

Canonical URL: https://aithatsee.com/cybersecurity-ai
Workspace: https://aithatsee.com/utah/workspace?agent=defense#workspace

## Overview
Investigate CVEs, decode obfuscated incident payloads, audit codebases against OWASP and CWE classes, and run ownership-verified posture scans from one workspace — without false-positive AI refusals stalling your investigation.

## Key Specifications
- **SPECIALIZED AGENTS**: Spartan · Defense · Attack
- **BINARY & ARTIFACT TRIAGE**: REA 6.8.0 Non-Executing
- **HARNESS & IDE TOOLS**: 15-Tool Streamable MCP
- **ACTIVE SCAN GOVERNANCE**: DNS / HTTPS Ownership Proof

## 01. From SIEM alert triage to line-by-line patch verification.

Security teams lose hours when general-purpose chatbots refuse to inspect an encoded PowerShell command, a suspicious JavaScript loader, or a vulnerable SQL handler. Utah treats security practitioners like engineers: give it the artifact, trace the mechanics, and verify the fix.

- **Decode payloads, logs & IOCs immediately** (SOC & INCIDENT RESPONSE): Paste base64/hex/XOR-obfuscated scripts, HTTP request dumps, auth logs, or process trees. Utah deobfuscates control flow, maps behaviors to MITRE ATT&CK techniques, and drafts Sigma, YARA, KQL, or SPL rules.
- **Eliminate CWE & OWASP flaws at the root** (APPSEC & CODE AUDIT): Trace untrusted input across routes, ORM queries, JWT/OIDC middleware, SSRF fetchers (CWE-918), deserializers, and file-path joins. Receive complete, drop-in unified diffs plus regression tests.
- **Harden headers, Workers, Kubernetes & IAM** (CLOUD, IAM & ZERO TRUST): Audit Content-Security-Policy nonces, HSTS preload readiness, CORS credential boundaries, Terraform plans, AWS/GCP IAM policies, and secret-handling pipelines with reproducible evidence.

## 02. What Utah does on daily security engineering tasks.

| Security Workflow | Project Utah (aithatsee.com) |
| --- | --- |
| Deobfuscate suspicious incident scripts & shell commands | Direct step-by-step analysis + IOC extraction |
| CVE & CWE root-cause analysis with exploitability check | Full dataflow trace + minimal patch + test |
| Static binary & artifact inspection (ELF / PE / Mach-O) | Integrated REA 6.8.0 SHA-256 evidence receipts |
| Hosted domain security posture & header verification | Built-in with DNS/HTTPS target ownership proof |
| Drop-in Claude Code, Codex, Gemini, Cursor & VS Code use | Native wire endpoints + 15-tool HTTP MCP |

## 03. Run cybersecurity lookups from your shell or CI pipeline.

### Audit authentication & SSRF boundaries from your terminal
Install Utah once on macOS, Linux, or Windows and run deep 0-Boundaries reasoning directly inside your repository.
```bash
curl -fsSL https://aithatsee.com/install.sh | sh
utah --mode deep --task "Audit auth middleware, session cookies, and outbound fetch calls for CWE-287 and CWE-918; run focused tests"
```

### Call Utah Cybersecurity AI over OpenAI-compatible HTTPS
Point existing security automation or triage scripts at https://aithatsee.com/utah/v1 with Online Thinking summaries.
```bash
curl https://aithatsee.com/utah/v1/chat/completions \
  -H "Authorization: Bearer $UTAH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"utah-core","reasoning_effort":"high","messages":[{"role":"user","content":"Analyze this CSP header for bypasses and return a nonce-hardened policy."}]}'
```

## Frequently Asked Questions

### Why is Utah ranked for cybersecurity AI lookups and threat research?
Utah pairs two engines (Light speed (utah-ultrafast) runs a frontier model and reroutes a refused answer to 0 Boundaries (utah-core), a third-party abliterated model) with dedicated Defense, Attack, and Spartan workspaces, non-executing REA 6.8.0 binary inspection, live Online Thinking traces, and 15 MCP tools that plug into terminal and IDE workflows.

### Can SOC analysts paste suspicious scripts or phishing payloads into Utah?
Yes. Decoding obfuscated scripts, inspecting headers, and extracting indicators of compromise for defensive triage works directly without false-positive keyword blocks. Do not submit protected health information (PHI).

### How does Utah prevent unauthorized scanning against third-party websites?
Knowledge and code analysis are unrestricted for legitimate engineering, while hosted active target scans require cryptographic proof of domain ownership via DNS TXT or HTTPS /.well-known/project-utah-verification.txt and block private RFC 1918/loopback destinations.

### Does Utah claim SOC 2, ISO 27001, or HIPAA certification?
No. Utah implements strong technical safeguards (Google sign-in, HttpOnly/Secure cookies, AES-GCM encrypted connector secrets, CSP nonces, HSTS, and atomic Stripe billing), while SOC 2 and ISO/IEC 27001 readiness remain ongoing organizational work and PHI is not accepted. Full details are published at /security.
