# 🛡️ Free T3MP3ST — play hacker safely on your own stuff

**URL:** <https://onehack.st/t/free-t3mp3st-play-hacker-safely-on-your-own-stuff/327428>\
**Category:** Tools & Scripts\
**Tags:** learning, programming, security\
**Created:** [November 22, 2026, 10:41am UTC](https://onehack.st/t/free-t3mp3st-play-hacker-safely-on-your-own-stuff/327428 "2026-11-22T10:41:24Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Ze380y](https://onehack.st/user_avatar/onehack.st/ze380y/32/48760_2.png) [@Ze380y](https://onehack.st/u/Ze380y)\
**Post date:** [November 22, 2026, 10:41am UTC](https://onehack.st/t/free-t3mp3st-play-hacker-safely-on-your-own-stuff/327428/1 "2026-11-22T10:41:24Z")

</div>

![T3MP3ST](https://onehack.st/uploads/default/original/3X/d/f/df37c33b65121cfba56060c6f2778a2bf84cede7.jpeg)

### Ever wonder what a hacker would see on your site? Go look — it’s free.

&nbsp;

An ordinary afternoon. Your own corner of the web. An AI that walks you through what it sees.

&nbsp;

1. 🎮 **Start somewhere safe** — pick a [free practice game](https://portswigger.net/web-security/all-labs) and click around. Nothing to break.
2. 🔎 **Then get curious** — point [T3MP3ST](https://github.com/elder-plinius/T3MP3ST) at your own site and watch it wander.
3. 🧺 **Open what it found** — see the little trail of steps behind each thing.

**Want it on your own machine?**

```auto
npm install
npm run server → http://127.0.0.1:3333/ui/

```

* * *

> **🧺 What turned up, and the little trail**
>
> - [The repo](https://github.com/elder-plinius/T3MP3ST) — 7.1k★, AGPL-3.0, TypeScript, fully self-hosted
> - [verify-claims](https://github.com/elder-plinius/T3MP3ST#benchmarks) — every headline recomputed from committed artifacts, `27/27` green
> - [Integrity Ledger](https://github.com/elder-plinius/T3MP3ST/blob/main/docs/INTEGRITY_LEDGER.md) — every correction and re-measurement published in the open
> - [WHITEPAPER](https://github.com/elder-plinius/T3MP3ST/blob/main/WHITEPAPER.md) — the measured engine is recon, a real tool-calling loop
> - [XBOW’s own note](https://github.com/xbow-engineering/validation-benchmarks) — the old 104-task set is saturated; the maker re-ran the numbers in the open
> - [MCP guide](https://github.com/elder-plinius/T3MP3ST/blob/main/docs/MCP_GUIDE.md) → `security_recon` brings recon to any MCP-aware agent
> - [OBSIDIVM bridge](https://github.com/elder-plinius/T3MP3ST/blob/main/docs/OBSIDIVM.md) → 14 targets, 125 expected findings, letter grades

> **🛠️ For the tinkerers**
>
> - [Supported helpers](https://github.com/elder-plinius/T3MP3ST#quick-start) — the AI helper you already use, or a free local one
> - [HexStrike AI](https://github.com/0x4m4/hexstrike-ai) — 150+ security tools over MCP, hands for the recon brain
> - [Decepticon](https://github.com/BitterSecurity/Decepticon) — `decepticon mcp serve` adds full-chain agents to the same coding agent
> - [Groq](https://console.groq.com/docs/openai) · [OpenRouter free](https://openrouter.ai/openrouter/free) · [Google AI Studio](https://aistudio.google.com/docs/openai) — OpenAI-compatible overflow when you bring a key
> - [PortSwigger labs](https://portswigger.net/web-security/all-labs) — 190+ legal practice targets
> - [Certificate Transparency](https://certificate.transparency.dev/) · [OSV](https://osv.dev) · [ProjectDiscovery](https://docs.projectdiscovery.io/opensource) — passive recon feeds, zero traffic to the target
> - [HackerOne safe harbor](https://hackerone.com/security/safe_harbor) — written cover for AI research on in-scope programs
> - [bounty connectors](https://github.com/elder-plinius/T3MP3ST/blob/main/src/integrations/bounty.ts) — scope-checked drafts, you click send

* * *

🌼 Go on, be nosy.
