Delete blackgat

blackgat/README.md

@@ -1,148 +0,0 @@
-# 🧠 Project BlackGat AI
-
-> _AI-powered Bug Bounty Automation Framework with Cybersecurity-Optimized LLMs_
-
-BlackGat is a modular, LLM-augmented bug bounty system designed to streamline vulnerability discovery, chaining, and report generation. It fuses the most advanced cybersecurity-focused language models into a unified tool for ethical hackers, red teamers, and security researchers.
-
----
-
-## 🚀 Key Features
-
-- 🧠 **LLM-Powered Agents**:
-  - **HeatSeeker** – Scores targets based on risk heuristics
-  - **KillChainAI** – Correlates vulnerabilities to suggest exploit chains
-  - **SCRIBE** – Autogenerates Markdown bug bounty reports
-  - **ReconGPT** – Converts natural language into scanning/fuzzing tasks
-  - **Hackphyr Agent** – Simulates attacker behavior, suggests exploits
-
-- 🤖 **Integrated LLMs from Hugging Face**:
-  | Model | Role |
-  |-------|------|
-  | `CyberBase-13B` | Report writing, CVE logic, PoC generation |
-  | `Llama-Primus-Reasoning` | Vuln correlation, chaining, logic mapping |
-  | `Hackphyr` | Red-team simulation, attacker tactics |
-  | `GPT-NeoX` | Prompt-driven recon and fuzzing suggestions |
-
-- 🧩 **Modular API with FastAPI**
-- 🖥️ **Live Dashboard using Streamlit**
-- ☁️ **Deployable on Hugging Face Spaces**
-- 🐳 **Docker-compatible**
-
----
-
-## 📦 Project Structure
-
-```
-blackgat/
-├── app/
-│   ├── main.py               # FastAPI API for AI endpoints
-│   ├── ai_modules.py         # All AI agent logic
-│   └── scanner_agent.py      # Reserved for future tool integrations
-├── models/
-│   └── load_llms.py          # LLM loader for Hugging Face models
-├── dashboard/
-│   └── dashboard.py          # Streamlit control panel
-├── requirements.txt          # Python dependencies
-├── Dockerfile                # Hugging Face & local deployment
-├── spaces.yaml               # Hugging Face Space config
-└── README.md                 # This file
-```
-
----
-
-## 🧠 Agents & Capabilities
-
-| Agent         | LLM Used                      | Capability                                                                 |
-|---------------|-------------------------------|----------------------------------------------------------------------------|
-| `HeatSeeker`  | Internal scoring              | Scores targets based on heuristic signals (e.g., admin pages, error codes) |
-| `KillChainAI` | `TrendMicro/llama-primus-reasoning` | Suggests chained vulnerabilities (e.g., IDOR + XSS → Account Takeover)   |
-| `SCRIBE`      | `CyberNative/CyberBase-13b`   | Writes vulnerability reports in Markdown format                           |
-| `ReconGPT`    | `EleutherAI/gpt-neox-20b`     | Converts natural language to recon/fuzzing actions                         |
-| `Exploit`     | `hackphyr/hackphyr`           | Suggests exploit paths and red-team logic                                 |
-
----
-
-## 🛠️ Installation (Local)
-
-```bash
-# Clone project
-git clone https://github.com/YOUR_USERNAME/blackgat.git
-cd blackgat
-
-# Install dependencies
-pip install -r requirements.txt
-
-# Run FastAPI backend (optional)
-uvicorn app.main:app --reload
-
-# Launch dashboard
-streamlit run dashboard/dashboard.py
-```
-
----
-
-## ☁️ Deploy to Hugging Face Spaces
-
-1. Go to [https://huggingface.co/spaces](https://huggingface.co/spaces)
-2. Create a new **Streamlit** space (call it `BlackGat-AI`)
-3. Clone it:
-
-```bash
-git lfs install
-git clone https://huggingface.co/spaces/YOUR_USERNAME/BlackGat-AI
-cd BlackGat-AI
-```
-
-4. Copy all files into the folder and push:
-
-```bash
-cp -r ../blackgat/* .
-git add .
-git commit -m "Initial BlackGat LLM AI deploy"
-git push
-```
-
-✅ Done. Your app will build and be live at:
-```
-https://huggingface.co/spaces/YOUR_USERNAME/BlackGat-AI
-```
-
----
-
-## ✅ Supported Use Cases
-
-- 🕵️‍♂️ Recon automation with LLM-driven payload suggestions
-- 🧠 Vulnerability classification and chaining
-- 📄 AI-assisted report generation (Markdown/JSON-ready)
-- ⚙️ Integration-ready for Burp Suite, sqlmap, ffuf, nuclei
-- 💬 Prompt-to-agent tasking for fuzzing, exploitation, and validation
-
----
-
-## 🧩 Roadmap (Next Phases)
-
-- [ ] Add `ffuf`, `sqlmap`, `dalfox` to the scanner agent
-- [ ] CVSS-based impact scoring
-- [ ] MongoDB backend for scan/result persistence
-- [ ] PDF report generation & auto-submission to HackerOne/Bugcrowd
-- [ ] User auth for multi-tenant dashboards
-
----
-
-## 🛡️ Ethical Usage
-
-BlackGat is intended for use in authorized security testing, research, and learning environments. Please ensure all activities conducted using this framework comply with legal and ethical standards. **Always respect target scope and obtain proper authorization.**
-
----
-
-## 🧠 Credits
-
-- Hugging Face – for hosting the cybersecurity-tuned LLMs
-- CyberNative, TrendMicro, Hackphyr, EleutherAI – for open-source LLMs
-- OWASP, HackerOne, ExploitDB – for vulnerability reference inspiration
-
----
-
-## 💬 License
-
-MIT License – Free to use, fork, and expand with attribution.

blackgat/app/ai_modules.py

@@ -1,52 +0,0 @@
-# app/ai_models.py
-import requests
-import os
-
-HF_API_TOKEN = os.getenv("BlackGat_AI")
-if not HF_API_TOKEN:
-    raise RuntimeError("BlackGat_AI token not set. Please add it as a Hugging Face Space secret.")
-
-HEADERS = {"Authorization": f"Bearer {HF_API_TOKEN}"}
-
-def query_huggingface(prompt, model_id, max_tokens=150):
-    url = f"https://api-inference.huggingface.co/models/{model_id}"
-    payload = {
-        "inputs": prompt,
-        "parameters": {"max_new_tokens": max_tokens}
-    }
-
-    try:
-        response = requests.post(url, headers=HEADERS, json=payload, timeout=30)
-        response.raise_for_status()
-        result = response.json()
-        if isinstance(result, list) and 'generated_text' in result[0]:
-            return result[0]['generated_text']
-        return "[ERROR] Unexpected response format"
-    except requests.exceptions.RequestException as e:
-        return f"[ERROR] API call failed: {str(e)}"
-
-def scribe_generate(data):
-    prompt = f"""Write a professional bug bounty report.
-Vulnerability: {data['type']}
-URL: {data['url']}
-Payload: {data['payload']}
-Impact: {data['impact']}"""
-    return {"report": query_huggingface(prompt, "schoolkithub/cyberbase", 200)}
-
-def killchain_ai(data):
-    prompt = f"Suggest a chained attack path using these findings: {data}"
-    return {"chain": query_huggingface(prompt, "schoolkithub/primus")}
-
-def heatseeker_score(data):
-    score = 0
-    if "admin" in data['url']: score += 3
-    if len(data.get("params", {})) > 3: score += 2
-    if data['status_code'] == 500: score += 5
-    return {"score": score}
-
-def recon_gpt(prompt):
-    return {"task": query_huggingface(prompt, "schoolkithub/neox")}
-
-def exploit_suggestion(data):
-    prompt = f"Given this bug bounty scenario, how might an attacker exploit it? {data}"
-    return {"exploit": query_huggingface(prompt, "schoolkithub/hackphyr")}

blackgat/app/main.py

@@ -1,51 +0,0 @@
-# app/main.py
-from fastapi import FastAPI
-from pydantic import BaseModel
-from typing import Dict
-from app.ai_models import (
-    scribe_generate, killchain_ai,
-    heatseeker_score, recon_gpt,
-    exploit_suggestion
-)
-
-app = FastAPI(title="BlackGat AI", version="1.0", description="Cloud AI for Bug Bounty Hunters")
-
-class VulnReport(BaseModel):
-    type: str
-    url: str
-    payload: str
-    impact: str
-
-class Finding(BaseModel):
-    data: str
-
-class ExploitRequest(BaseModel):
-    data: str
-
-class HeatInput(BaseModel):
-    url: str
-    params: Dict[str, str]
-    status_code: int
-
-class ReconPrompt(BaseModel):
-    prompt: str
-
-@app.post("/scribe")
-def scribe(data: VulnReport):
-    return scribe_generate(data.dict())
-
-@app.post("/killchain")
-def killchain(data: Finding):
-    return killchain_ai(data.data)
-
-@app.post("/score")
-def score(data: HeatInput):
-    return heatseeker_score(data.dict())
-
-@app.post("/recon")
-def recon(data: ReconPrompt):
-    return recon_gpt(data.prompt)
-
-@app.post("/exploit")
-def exploit(data: ExploitRequest):
-    return exploit_suggestion(data.data)

blackgat/dashboard/dashboard.py

@@ -1,59 +0,0 @@
-# dashboard/app.py
-import streamlit as st
-import requests
-
-API_URL = "http://localhost:8000"
-
-st.set_page_config(page_title="BlackGat AI", layout="wide")
-st.title("🧠 BlackGat AI — Bug Bounty Automation")
-
-tool = st.sidebar.selectbox("Choose Agent", ["Scribe", "KillChain", "HeatSeeker", "ReconGPT", "Exploit Suggestion"])
-
-if tool == "Scribe":
-    st.subheader("📄 Generate Bug Bounty Report")
-    vuln_type = st.text_input("Vulnerability Type")
-    vuln_url = st.text_input("Vulnerable URL")
-    payload = st.text_input("Payload")
-    impact = st.text_area("Impact")
-    if st.button("Generate"):
-        res = requests.post(f"{API_URL}/scribe", json={
-            "type": vuln_type,
-            "url": vuln_url,
-            "payload": payload,
-            "impact": impact
-        })
-        st.markdown(res.json()["report"])
-
-elif tool == "KillChain":
-    st.subheader("🔗 Suggest Chained Attack")
-    findings = st.text_area("Paste findings (JSON or summary)")
-    if st.button("Suggest"):
-        res = requests.post(f"{API_URL}/killchain", json={"data": findings})
-        st.markdown(res.json()["chain"])
-
-elif tool == "HeatSeeker":
-    st.subheader("🌡️ Risk Scoring")
-    url = st.text_input("Target URL")
-    status_code = st.number_input("Status Code", min_value=100, max_value=599, value=200)
-    params = st.text_input("Params JSON", '{"user":"admin"}')
-    if st.button("Score"):
-        res = requests.post(f"{API_URL}/score", json={
-            "url": url,
-            "params": eval(params),
-            "status_code": status_code
-        })
-        st.success(f"Score: {res.json()['score']}")
-
-elif tool == "ReconGPT":
-    st.subheader("🛰️ Recon Task Generator")
-    prompt = st.text_area("What do you want to find?", "Find all login endpoints.")
-    if st.button("Generate Task"):
-        res = requests.post(f"{API_URL}/recon", json={"prompt": prompt})
-        st.markdown(res.json()["task"])
-
-elif tool == "Exploit Suggestion":
-    st.subheader("💥 Attacker Simulation")
-    scenario = st.text_area("Bug scenario or data")
-    if st.button("Get Exploit Advice"):
-        res = requests.post(f"{API_URL}/exploit", json={"data": scenario})
-        st.markdown(res.json()["exploit"])

blackgat/models/load_llms.py

@@ -1,21 +0,0 @@
-from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
-
-def load_llms():
-    return {
-        "cyberbase": pipeline("text-generation",
-            model=AutoModelForCausalLM.from_pretrained("CyberNative/CyberBase-13b", device_map="auto"),
-            tokenizer=AutoTokenizer.from_pretrained("CyberNative/CyberBase-13b")
-        ),
-        "primus": pipeline("text-generation",
-            model=AutoModelForCausalLM.from_pretrained("TrendMicro/llama-primus-reasoning", device_map="auto"),
-            tokenizer=AutoTokenizer.from_pretrained("TrendMicro/llama-primus-reasoning")
-        ),
-        "hackphyr": pipeline("text-generation",
-            model=AutoModelForCausalLM.from_pretrained("hackphyr/hackphyr", device_map="auto"),
-            tokenizer=AutoTokenizer.from_pretrained("hackphyr/hackphyr")
-        ),
-        "neox": pipeline("text-generation",
-            model=AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neox-20b", device_map="auto"),
-            tokenizer=AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
-        )
-    }

blackgat/requirements.txt

@@ -1,8 +0,0 @@
-fastapi
-uvicorn
-streamlit
-transformers
-torch
-requests
-pydantic
-Gradio

blackgat/spaces.yaml

@@ -1,2 +0,0 @@
-sdk: streamlit
-app_file: dashboard/dashboard.py