S1N6H/pentest-harness

S1N6H/pentest-harness has 461 GitHub stars and 80 forks. GitHot measured -1 stars in 24 hours. Last checked 2026-11-10 04:39 UTC. Last push 2026-10-07.

How these counts are measured

Pentest Harness — Heaven for Hackers. A self-hosted AI agent harness for authorized pentests, bug bounty, security labs, and CTFs. Bring your own AI model API; sessions stay local.

461 stars 80 forks

What you can do

Multi-provider LLM engine — OpenAI Chat Completions & Responses, Anthropic Messages, DeepSeek, and any custom OpenAI-compatible gateway with one-click model auto-discovery
Full agent toolset — shell, filesystem, web research, skills, goals, subagents, background jobs, and workflow control
Pentest Mode — a professional offensive-security operating standard for authorized engagements
Durable sessions — JSONL/SQLite persistence with replay; resume exactly where you left off

Measured star growth

in 24h: -1
in 7 days: Collecting history
in 15 days: Collecting history
in 1 month: Collecting history

Net star changes over overlapping periods, not daily totals. Missing measurements are left blank. Stars measure attention, not software quality.

README.md

Original

Pentest Harness

Heaven for Hackers — the professional security testing workspace.

An open-source, dark-first AI agent harness for authorized penetration tests, bug bounty research, security labs, and CTF engagements. Built on a plugin architecture where every layer — model adapters, tools, sessions, settings, and credentials — is replaceable from configuration.

Works with any AI model API. Bring your own key from OpenAI, Anthropic, DeepSeek, Google, Mistral, Groq, OpenRouter, Azure OpenAI, or any OpenAI-compatible gateway — one click auto-discovers your models and you're running.

Screenshots

Workspace	Provider cards

Features

Multi-provider LLM engine — OpenAI Chat Completions & Responses, Anthropic Messages, DeepSeek, and any custom OpenAI-compatible gateway with one-click model auto-discovery
Full agent toolset — shell, filesystem, web research, skills, goals, subagents, background jobs, and workflow control
Pentest Mode — a professional offensive-security operating standard for authorized engagements
Provider cards — live connection testing, enabled/disabled toggles, per-model context badges
Durable sessions — JSONL/SQLite persistence with replay; resume exactly where you left off
Context that never dies — token metering, automatic compaction, and tool-result pruning
Secure by default — API keys live in an owner-only credential store as references; never in settings files or logs
Dark-only theme — a focused, low-glare interface built for long engagements
Everything is a plugin — built on Cordis; no closed core to fight

Demo

Download the full video

Quick start

git clone https://github.com/S1N6H/pentest-harness.git
cd pentest-harness
pnpm install
pnpm build
pnpm dsh web

The Web UI opens at http://127.0.0.1:2323 in your browser.

If pnpm dsh is not on your PATH, use pnpm exec dsh web or node apps/cli/lib/index.js web.

Requirements

Node.js 20+ (Node 22 recommended)
pnpm (npm i -g pnpm)

Run on another port

pnpm dsh web --port 3000

First run — connect any AI model

Open Settings → Models.
Click Add a custom provider (or Edit an existing one).
Paste your API base URL — the provider ID and display name auto-fill.
Enter your API key (stored in your local credential store, never in settings).
Models auto-discover from the endpoint — click one to add it, or add by hand.
Pick a model in the composer and start your engagement.

Defaults

Setting	Value
Web UI	`http://127.0.0.1:2323`
Settings file	`$DSH_HOME/settings.yaml` (default `~/.dsh/settings.yaml`)
Credentials	`$DSH_HOME/.credentials.yaml` (owner-only permissions)

License

MIT

⚡ Sponsored by HackWithClaude

Unlimited AI models · One platform · Zero limits

Access every major model — Claude, GPT, Gemini, and more — in one place. Built for builders who ship.

🚀 Visit HackWithClaude.com · 💡 Explore the models