Name: 0sec
Author: 0sec-labs

0sec

Your open-source AI cybersecurity agent.
It hacks, proves the problem, and writes the fix. Multi-model, multi-agent, but most importantly: yours.

_{🇨🇭 Swiss Applied AI Cybersecurity Research Lab · 0.security}

release

_{Beta — in active development; interfaces may change. See Honest limitations.}

0sec — finds, proves, and reports vulnerabilities

Install

curl -fsSL https://raw.githubusercontent.com/0sec-labs/0sec/main/install.sh | bash
0 --help

Verified release binary (SHA-256 checked) → ~/.0sec/bin. macOS (Apple Silicon) and Linux (x64/arm64); Windows: 0sec-windows-x64.exe from the release; Docker: ghcr.io/0sec-labs/0sec:latest. Not on npm.

Quick start

# 1. say what you're allowed to touch
echo '{ "in_scope": ["example.com"] }' > scope.json
export ANTHROPIC_API_KEY=...        # or OpenAI, Azure, OpenRouter, Ollama, …

# 2. scan a live target (out-of-scope requests are refused)
0 scan --target https://example.com --scope ./scope.json

# 3. or review code, audit a package, open the console
0 review ./my-app                   # source review
0 audit lodash                      # npm / pypi / cargo / oci package
0 console --scope ./scope.json      # interactive; type / for commands

0sec quickstart — a scan from start to finish

What it covers

Most tools stop at the app. 0sec goes all the way down.

Layer	Finds
Web apps	SQLi, IDOR, XSS, SSRF, auth bypass
APIs	tenant isolation, BOLA, business-logic abuse
AI & LLMs	prompt injection, jailbreaks, MCP tool abuse
Source code	injection, auth, deserialization, memory safety
Dependencies	supply chain, malicious packages, CVE replay
Network / identity	AD, cloud, federation (read-only, offline)
Runtime / OS / kernel	container escape, privesc, 0-day hunt
Compiled binaries	no source → `0verse`

Commands

Task	Commands
Pentest web / AI-LLM / MCP	`scan`, `eval`, `agent-assure`
Review source / packages / kernel	`review`, `file-review`, `deep-review`, `audit`
Recon an attack surface	`recon`, `js-recon`, `npm-discovery`, `intel`
Hunt a bug class / kernel variants	`hunt`, `kernel`, `cve`
Work with evidence	`findings`, `history`, `resume`, `replay`, `verify`, `disclose`
Generate & re-test a fix	`fix`
Identity / AD (read-only)	`identity`, `adgraph`, `entragraph`
Integrate	`mcp-server`, `console`, `tui`, `dashboard`

Run 0 --help for the rest. Full docs: docs.0.security.

0sec console command palette
_{The interactive console — / opens the command palette.}

How it works

It proves the bug before it reports it.

Free-form agents, hard guardrails. Models decide what to probe; turn budgets, loop detection, and scope-on-every-call keep them in line.
Reproduce before trust. A blind agent re-exploits each finding from the PoC alone. What it can't reproduce is dropped.
Triage before verify. Class oracles and a second scanner cut noise before the expensive step.
Bring your own model. Anthropic, OpenAI, Azure, OpenRouter, or local Ollama — you hold the key.

Every run keeps its own evidence under ~/.0sec/runs/<id>/, so you can resume, replay, or disclose it later.

0sec blind verification
_{Blind verification — every finding is re-exploited before it ships.}

Track record

0sec has landed real, maintainer-reviewed fixes in the mainline Linux kernel and other open source. The verified list lives at 0.security. Benchmarks are secondary evidence — caveats in the benchmark docs.

Supported by

With special thanks to the startup and research programs supporting our work:

Honest limitations

Kernel/IOKit findings stay hypotheses until a real oracle reproduces them (the linux-kernel profile is static).
Verification depth varies: verificationSpec covers file/diff predicates; the replay runner is local-shell only (Docker/QEMU are stubs).
The false-positive-moat layers are off by default and slice-dependent.
Benchmarks are single-model/config/trial; the 10/10 AI-suite is self-authored, not independent.
fix is narrow: source-only, single-file, ≤3 attempts.
By design, never: network sweeps, credential spraying, persistence/C2, or stealth.

Build from source

git clone https://github.com/0sec-labs/0sec.git && cd 0sec
corepack enable && pnpm install --frozen-lockfile && pnpm build && node dist/0sec.js --help

Contributing & security

See CONTRIBUTING.md — synthetic or authorized targets only. Report vulnerabilities privately via SECURITY.md ([email protected]), not public issues.