Public Beta

Autonomous Offensive Security

AI agents that recon, exploit, and validate — like having a red team on autopilot. Adopted from Strix, hardened for Indonesian infrastructure. No manual intervention, no hand-holding.

"Secepat petir menyambar, seakurat itu pula kami menemukan celahmu."

50+
Tools in sandbox
40+
Exploit skills
3
Attack modes
0
API keys sold

Runs on

OpenAI Anthropic Ollama Docker Kali

What the agents actually do

Dari enumeration sampai validasi PoC. Semuanya dieksekusi di dalam sandbox, bukan simulasi.

🤖

Multi-Agent Orchestration

Agent pool menjalankan recon, mapping, exploitation, dan reporting secara paralel. Satu target, banyak vektor.

🔧

Real Tools, Real Output

nmap, nuclei, sqlmap, semgrep, gitleaks, Caido, headless Chromium — dijalankan langsung, bukan mock.

🎯

Anything as Target

Live URLs, repositori Git, kode lokal, OpenAPI specs, Postman collections. Petir akan mencari vektor dari situ.

📊

Live Dashboard

Pantau scan secara real-time. Severity breakdown, attack chain visualizer, per-agent progress. Bukan log file.

📝

Compliance Reports

Output dalam SARIF, CVSS 4.0, executive summary. Ada template UU ITE, UU PDP, dan ISO 27001.

🔄

Pipeline Native

Jalankan di GitHub Actions, GitLab CI, Bitbucket Pipelines. Block deployment kalo finding critical muncul.

🧠

Domain-Specific Skills

SQLi, XSS, RCE, SSRF, JWT attacks, OAuth misconfig, prototype pollution, deserialization — bukan generic scan.

🇮🇩

Indonesian-First

UI, reporting, dan knowledge base dalam Bahasa Indonesia. Dibuat untuk tim yang kerja di infrastruktur lokal.

Tiga level kedalaman scan

Dari sanity check sampai full-blown adversarial simulation.

Quick

5–15m

High-signal, low-noise. Langsung ke OWASP Top 10 dan CWE Top 25. Cocok buat pre-commit hook dan rapid triage.

🎯

Standard

30–60m

Full attack surface enumeration, authenticated crawling, multi-stage exploit chain. Default mode untuk CI/CD pipeline.

🔬

Deep

1–4h

Exhaustive. Business logic abuse, race conditions, chained exploits, lateral movement simulation. Production audit grade.

How it runs under the hood

Modular, containerized, dan tidak menyentuh host Anda secara langsung.

Web Dashboard (React)
Projects  |  Scans  |  Findings  |  Reports  |  Team
REST API (FastAPI)
Auth  |  Scan Queue  |  Scheduler  |  Webhooks
Engine (Python 3.12 + Go TUI)
Agent Orchestrator → Sandbox Container (Kali + tools)
Persistence Layer
PostgreSQL  |  Redis  |  S3-compatible  |  Kubernetes

Built for people who ship code

Bukan sekadar scanner. Ini alat buat tim yang mau security diintegrasikan ke workflow, bukan di-check setelah production.

👨‍💻

Engineers

Scan dari localhost sebelum push. Petir jadi bagian dari dev loop, bukan afterthought.

🛡️

AppSec Teams

Laporan compliance otomatis. Template SARIF, CVSS, UU ITE, UU PDP, ISO 27001 — tinggal export.

🚀

Startup & UMKM

Pentest tanpa harus hire konsultan atau beli lisensi enterprise. Apache 2.0, self-hosted.

🎯

Bug Bounty

Otomasi reconnaissance dan triage. Fokus waktu kamu di validasi dan report writing.

The components

Dipilih untuk stabilitas dan performa, bukan hype.

Python 3.12+
Core
FastAPI
API Layer
PostgreSQL
Database
React + Vite
Frontend
Tailwind CSS
Styling
Docker + Kali
Sandbox
Celery + Redis
Queue
Go + Bubble Tea
TUI
LiteLLM
LLM Router
OpenTelemetry
Observability

One command to start scanning

Install, set provider, run. Tidak ada konfigurasi YAML 300 baris.

shell
# Install
curl -sSL https://petir.dev/install | bash

# Pilih LLM provider
export PETIR_LLM="openai/gpt-4o"            # OpenAI
export PETIR_LLM="anthropic/claude-sonnet-4-6" # Anthropic
export PETIR_LLM="ollama/llama3"            # Self-hosted via Ollama

export LLM_API_KEY="sk-..."

# Scan
petir --target https://example.com
petir --target ./src              # Static analysis pada kode
petir --target gh:user/repo      # Clone dan scan repository

Find the holes before they do.

Open source. Apache 2.0. Self-hosted atau cloud. Terserah kamu.