More screenshots
Works with the AI Agents you already use
omni maximus
Autonomous Adversarial SRE and AppSec engine for rigorous code verification and zero-false-positive vulnerability PoCs.
$39
omni maximus
Example session with this skill installed
Audit the user registration flow in /src/auth/register.ts for concurrency issues and physical limits.
- Read your context and instructions
- Compiled the omni maximus
I have completed a 6-Gate Adversarial Audit of the registration flow. A critical SEV-2 concurrency vulnerability was identified regarding duplicate wallet creation during rapid request bursts.
Adversarial Audit Report
<supply_chain_gate>
<manifests status="VERIFIED" lockfile_present="true" />
<secrets_risk detected="NONE" />
<sentinel_violations count="0" />
</supply_chain_gate>
| ID | Type | Route / Symbol | File:Line | Auth Level | Reached Data Sinks |
|---|---|---|---|---|---|
| EP-01 | POST | /api/register | register.ts:12 | Public | Database:Users, Database:Wallets |
<taint_matrix>
<vector id="SEV-2">
<source>Concurrent HTTP POST requests with identical email/identity.</source>
<flow>validate() -> createUser() -> createInitialWallet()</flow>
<sink file="src/auth/register.ts" line="42">Wallet table insert lacks row-level lock or atomic check.</sink>
</vector>
</taint_matrix>
<falsation_audit>
- Taint / Stress Check: Confirmed. The wallet creation occurs after the user insert but before the transaction commits without a mutex.
- Executable PoC Vector:
for i in {1..5}; do curl -X POST http://localhost:3000/api/register -d '{"email":"test@example.com"}' & done; - Runtime Lifecycle: Standard Node.js cluster mode will allow race condition.
- Verdict: CONFIRMED.
</falsation_audit>
// Phase 1: Atomic Concurrency Fix
async function registerUser(data: RegData) {
return await db.transaction(async (tx) => {
// Added SELECT FOR UPDATE to lock identity row
const existing = await tx.select().from(users).where(eq(users.email, data.email)).for('update');
if (existing) throw new Error("Conflict");
const newUser = await tx.insert(users).values(data).returning();
await tx.insert(wallets).values({ userId: newUser.id, balance: 0 });
return newUser;
});
}
Next steps
- Run the provided shell script to verify the fix prevents duplicate wallets.
- Review the
LIMITclauses on the user lookup to satisfy SEV-5 requirements. - Convene the 4-Voice Council to discuss migrating to a distributed lock if scaling horizontally.
Connects securely to your tools. The creator never sees your data.
About this skill
Why Standard AI Coding Agents Fail in Production
AI coding models suffer from conversational sycophancy and self-referential blindness: when asked to review their own code, they say "everything looks clean and well-structured" while missing critical bugs:
- Unbounded database queries without
LIMITthat exhaust connection pools under load. - Concurrency race conditions in balances and inventory permitting double-spending.
- UI state machines where sequential handlers silently cancel each other out.
- Handcrafted regexes vulnerable to Catastrophic Backtracking (ReDoS).
OMNI-MAXIMUS transforms your AI into a ruthless, zero-sycophancy Adversarial Red-Team.
🛡️ The 6 Hard Gates Pipeline
- Gate 0 (Supply Chain & Sentinel NIH Shield): Enforces committed lockfiles, scans for hardcoded secrets, and blocks handcrafted crypto or homebrew JWT parsers.
- Gate 1 (AST Surface Inventory): Maps all public routes, webhooks, and workers with their call paths to critical data sinks.
- Gate 2 (SRE Physical Limits): Enforces computational limits: heap streams (<= 8 KB), CAS atomic locks, and mandatory
LIMITclauses on all collection queries. - Gate 3 (UI Forensics & Click-Path Audit): Detects silent state cancellations in React/Zustand/Redux, WCAG 2.2 AA contrast (4.5:1), and zero horizontal scroll on 360px.
- Gate 4 (Mandatory PoC Falsification): Zero False Positives: Every reported issue MUST synthesize an executable
curlcommand or concurrency script proving the failure. - Gate 5 (Surgical 2-Phase Remediation): Applies atomic diffs verified by dual independent reviewers (Santa Method) requiring a resilience score >= 95/100.
⚡ Universal Compatibility
Fully compliant with the Open SKILL.md Standard. Works natively across Claude Code, Cursor, Windsurf, Google Antigravity, Copilot, and Amazon Q.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Requires lockfiles for supply chain checks. PoC execution depends on the agent having terminal access. Scope is limited to code and architectural forensics, not live infrastructure monitoring.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
Permissions
Allowed Hosts