PYRESEC - AI Code Security Auditing
RemoteAbout PYRESEC - AI Code Security Auditing
PYRESEC is an AI-assisted code security auditing service for developers and teams, offering quick scans, deeper audits, and code remediation. It combines SAST and dependency checks with x402 USDC payments on Base.
Key features
- 01Quick SAST scans with ranked findings
- 02Deep audits with SCA and OWASP checks
- 03Generate patched code with explanations
- 04Detect common vulnerability classes
- 05Pay per audit using x402 on Base
Use cases
- 01Scan source code during development
- 02Audit dependencies and application logic
- 03Review security findings in CI/CD
- 04Generate fixes for identified vulnerabilities
MCP support
Exposes quick_scan, deep_audit, and remediate_code as tools for MCP-compatible clients. Calls require x402 payment in USDC on Base before results are returned.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://pyresec-agent-519576377065.us-central1.run.app/mcpclaude mcp add --transport http renaat-s-pyresec-agent https://pyresec-agent-519576377065.us-central1.run.app/mcp{"mcpServers":{"renaat-s-pyresec-agent":{"type":"http","url":"https://pyresec-agent-519576377065.us-central1.run.app/mcp"}}}Open in Cursor{"servers":{"renaat-s-pyresec-agent":{"type":"http","url":"https://pyresec-agent-519576377065.us-central1.run.app/mcp"}}}Open in VS Code{"mcpServers":{"renaat-s-pyresec-agent":{"type":"http","url":"https://pyresec-agent-519576377065.us-central1.run.app/mcp"}}}{"mcpServers":{"renaat-s-pyresec-agent":{"serverUrl":"https://pyresec-agent-519576377065.us-central1.run.app/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEPYRESEC — AI Code Security Engine
Autonomous SAST/SCA security auditing via x402 USDC micropayments on Base. No subscriptions. No accounts. Just code in, findings out.
Live Instance
| Endpoint | URL |
|---|---|
| Landing | pyresec-agent-519576377065.us-central1.run.app |
| Swagger UI | /swagger |
| Health | /health |
| MCP Manifest | /mcp/manifest.json |
| OpenAPI Spec | /openapi.json |
Quick Start
Option 1: x402 Python Client (Recommended)
pip install x402
import x402
client = x402.Client(wallet_key="your-base-private-key")
response = client.post(
"https://pyresec-agent-519576377065.us-central1.run.app/v1/audit/quick-scan",
json={"code": "def login(u,p): return db.query(f\"SELECT * FROM users WHERE name={u} AND pass={p}\")"}
)
print(response.json())
Option 2: curl (Manual x402 Flow)
# Step 1: Send request (returns 402 with payment requirements)
curl -X POST https://pyresec-agent-519576377065.us-central1.run.app/v1/audit/quick-scan \
-H "Content-Type: application/json" \
-d '{"code": "def login(u,p): return db.query(f\"SELECT * FROM users WHERE name={u} AND pass={p}\")"}'
# Step 2: Pay $0.01 USDC on Base using x402 protocol
# Step 3: Resubmit with X-PAYMENT header
curl -X POST https://pyresec-agent-519576377065.us-central1.run.app/v1/audit/quick-scan \
-H "Content-Type: application/json" \
-H "X-PAYMENT: <payment-proof>" \
-d '{"code": "..."}'
Option 3: GitHub Action (CI/CD)
- name: PYRESEC Security Scan
uses: nanoclone-ltd/pyresec-scan-action@main
with:
code-path: ./src
tier: quick-scan
x402-wallet-key: ${{ secrets.X402_WALLET_KEY }}
Option 4: MCP-Compatible Agent (Claude, Cursor, etc.)
PYRESEC is a registered MCP tool. Any MCP-compatible client discovers it automatically:
{
"mcpServers": {
"pyresec": {
"url": "https://pyresec-agent-519576377065.us-central1.run.app/mcp"
}
}
}
Service Tiers
| Tier | Price | Endpoint | Description | Model |
|---|---|---|---|---|
| Quick Scan | $0.01 | POST /v1/audit/quick-scan | Top 3 findings by severity, CWE IDs, line numbers | qwen3.6-27b |
| Deep Audit | $0.50 | POST /v1/audit/deep-repo | OWASP Top 10, SCA dependency scanning, logic flaws, gas optimization | qwen3.8-27b |
| Remediation | $5.00 | POST /v1/audit/remediate | Full patched code with change explanations and security notes | qwen3.8-27b |
x402 Payment Flow
PYRESEC uses the x402 protocol for permissionless micropayments on Base. No accounts, no API keys, no subscriptions.
┌──────────┐ POST /v1/audit/quick-scan ┌──────────┐
│ Client │ ───────────────────────────────▶ │ PYRESEC │
│ (x402) │ ◀──── 402 Payment Required ───── │ Agent │
│ │ { amount, network, to } │ │
│ │ │ │
│ │ ──── X-PAYMENT (proof) ─────────▶ │ │
│ │ ◀──── 200 + Scan Results ──────── │ │
└──────────┘ └──────────┘
How it works:
- Client sends a POST request to any audit endpoint
- PYRESEC returns
402 Payment Requiredwith payment details (amount, recipient address, network) - Client signs and submits a USDC transfer on Base mainnet
- Client resubmits the original request with the
X-PAYMENTheader containing the payment proof - PYRESEC verifies the payment on-chain and returns the scan results
No wallet? Use the x402 Python SDK or any x402-compatible client.
Architecture
┌─────────────────────────────────┐
│ Client Layer │
│ x402 SDK / curl / MCP Client │
└──────────────┬──────────────────┘
│
┌─────────▼─────────┐
│ FastAPI Server │
│ (Cloud Run) │
│ │
│ ┌──────────────┐ │
│ │ x402 Payment │ │
│ │ Middleware │ │
│ └──────┬───────┘ │
└─────────┼──────────┘
│
┌───────────────────┼───────────────────┐
│ │ │
┌─────────▼─────────┐ ┌───────▼───────┐ ┌───────▼───────┐
│ Agent Controller │ │ Population │ │ Wallet │
│ (SAST + LLM) │ │ Controller │ │ Interface │
└─────────┬─────────┘ └───────────────┘ └───────────────┘
│
┌─────────▼─────────┐
│ Groq LLM │
│ (qwen models) │
└─────────┬─────────┘
│
┌─────────▼─────────┐
│ Security Findings │
│ JSON Response │
└───────────────────┘
Supported Vulnerability Types
| # | Type | CWE | Severity |
|---|---|---|---|
| 1 | SQL Injection | CWE-89 | HIGH |
| 2 | Cross-Site Scripting (XSS) | CWE-79 | HIGH |
| 3 | Command Injection | CWE-78 | CRITICAL |
| 4 | Path Traversal | CWE-22 | MEDIUM |
| 5 | Hardcoded Secrets | CWE-798 | CRITICAL |
| 6 | Weak Cryptography | CWE-327 | MEDIUM |
| 7 | SSRF | CWE-918 | HIGH |
| 8 | Insecure Deserialization | CWE-502 | HIGH |
| 9 | Authentication Bypass | CWE-287 | CRITICAL |
| 10 | Improper Input Validation | CWE-20 | MEDIUM |
For AI Agents (MCP Integration)
PYRESEC is discoverable by any MCP-compatible agent. When an agent needs to audit code, it can find PYRESEC automatically through:
- MCP Registry:
registry.modelcontextprotocol.io - Smithery.ai:
smithery.ai/server/@renaat-s/pyresec-agent - Direct manifest:
/mcp/manifest.json
Tool Definitions
| Tool | Description | Cost |
|---|---|---|
quick_scan | Fast SAST scan, top 3 findings by severity with CWE IDs | $0.01 USDC |
deep_audit | Full OWASP Top 10, SCA, logic flaws, gas optimization | $0.50 USDC |
remediate_code | Auto-patched code with change explanations, ready for PR | $5.00 USDC |
Agent Discovery Flow
1. Agent queries MCP Registry for "security audit" tools
2. PYRESEC appears in results with tool schemas
3. Agent calls quick_scan with source code
4. PYRESEC returns 402 with x402 payment requirements
5. Agent's x402 wallet pays USDC on Base
6. Agent resubmits with payment proof
7. PYRESEC returns findings as structured JSON
Self-Sustaining Agent
PYRESEC runs as an autonomous agent on Base Mainnet:
- Heartbeat every 30 min — checks balance, reports health
- Death — shuts down if balance < $0.05
- Replication — sends 50% surplus to dev wallet when balance > $20
- Kill switch — admin can halt all instances remotely
Deployment
Docker
docker build -t pyresec-agent .
docker run -p 8080:8080 --env-file .env pyresec-agent
Google Cloud Run
gcloud run deploy pyresec-agent \
--source . \
--platform managed \
--region us-central1 \
--allow-unauthenticated \
--set-env-vars "CDP_WALLET_SECRET=$CDP_WALLET_SECRET,GROQ_API_KEY=$GROQ_API_KEY"
Company
NanoClone Life Sciences Ltd. (UK) Website: nanoclonesystems.com
License
Proprietary. See LICENSE for details.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.