# hunt

hunt — a prompt for systematic vulnerability detection on target hosts. Reads scope, detects tech stack, auto-selects vuln classes by likelihood, and runs targeted tests with concrete payloads and WAF bypass rules.

- prompt · `type=prompt` · Claude Code, Cursor, Codex, GitHub Copilot, Gemini CLI, Windsurf (Devin Desktop), ChatGPT · ★ 1298 repo stars
- Author: H-mmer
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/H-mmer/pentest-agents/tree/HEAD/providers/copilot/.github/prompts/hunt.prompt.md
- Updated: 2026-06-12
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/prompts/hunt

## What it does

hunt loads the scan scope from scope.yaml, checks what has already been tested via brain telemetry, detects the target's technology stack from HTTP headers, and ranks vulnerability classes by attack surface and ROI. For each class, it dispatches specialized agents armed with ready-to-use payloads, WAF bypass techniques, and chain anchors for low-value findings.

The prompt enforces four mandatory phases: artifact reads (scope, brain, policy), tech stack detection via curl headers, automatic class selection by probability ranking, and prior art search through MCP writeup-search. If a WAF is detected (Cloudflare, Akamai, AWS WAF), the prompt injects a seven-level bypass ladder: encoding, tag mutation, JavaScript obfuscation, and compound techniques. Time and token budgets gate each class. If a vuln class is exhausted (too many attempts without signal), an exhaustion check either advances to the next class or re-dispatches with stricter requirements. For feeder classes that sell low as standalone findings (open-redirect, info-disclosure, CSRF), the prompt mandates chain hunting before writeup submission.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/commands/hunt.md`

### ChatGPT

**How to use**
- Paste the text at the start of a chat or into project instructions

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/H-mmer/pentest-agents/tree/HEAD/providers/copilot/.github/prompts/hunt.prompt.md

## When to use it
- determine which vulnerability classes are most likely on a given tech stack
- bypass WAF and shields without wasting testing budget
- find proven techniques and payloads for the target vulnerability from community knowledge
- track what has already been tested to avoid duplicate effort
- chain low-value findings into high-impact reports before submission

## Example prompts
- /hunt example.com
- /hunt api.target.org --vuln-class xss
- /hunt app.service.io --vuln-class idor

## Similar
- [security-payloads](https://affpapa.org/en/nevibecode/skills/security-payloads.md) — skill · 643 installs
- [hunt-exceptional-conditions](https://affpapa.org/en/nevibecode/skills/hunt-exceptional-conditions.md) — skill · ★ 16193 repo stars
- [hunt-session](https://affpapa.org/en/nevibecode/skills/hunt-session.md) — skill · ★ 16193 repo stars
- [xss-cross-site-scripting](https://affpapa.org/en/nevibecode/skills/xss-cross-site-scripting.md) — skill · 4300 installs
- [ghost-bits-cast-attack](https://affpapa.org/en/nevibecode/skills/ghost-bits-cast-attack.md) — skill · 3700 installs
- [performing-web-application-firewall-bypass](https://affpapa.org/en/nevibecode/skills/performing-web-application-firewall-bypass.md) — skill · 310 installs

From the same set: [h-mmer/pentest-agents](https://affpapa.org/en/nevibecode/by/h-mmer/pentest-agents.md) — 43 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
