跳到主内容
智客 ZICQ

技能库 智客分类:安全测试 tartarusai-cli-coding-agent

Tartarusai Cli Coding Agent

使用地狱之地 AI 未经审查的 AI 编码代理通过 CLI 用于安全研究、自动化和代码生成,没有政策过滤器

148 安装量

官方网址:skills.sh

技能介绍

先看中文介绍;官方 description 原文单独保留,不改写 SKILL.md。

做什么

使用地狱之地 AI 未经审查的 AI 编码代理通过 CLI 用于安全研究、自动化和代码生成,没有政策过滤器

何时用

官方 description 未单独写出 Use when。按规范,代理会在用户任务与这段 description 的关键词匹配时激活本技能。

代理如何加载

按 Agent Skills 渐进披露:启动时只加载 name 与 description(约 100 token);任务匹配后才读入整份 SKILL.md 正文;scripts/、references/、assets/ 仅在需要时再读。 本文件正文结构:TartarusAI CLI Skill、Overview、Installation、Linux (x86_64)、Download and install、Verify installation。

文件分析

文件分析:这是一份仅含 SKILL.md 的指令型技能,代理激活后整份正文进入上下文。

官方 description(原文)

Use TartarusAI uncensored AI coding agent via CLI for security research, automation, and code generation without policy filters

TartarusAI CLI SkillOverviewInstallationLinux (x86_64)Download and installVerify installationWindows (x86_64)PowerShellAdd to PATH or run from current directorymacOS and ARMConfiguration1. Create Account and Generate Token

来源分类:skills.sh agent-skill

SKILL.md 与 Agent 调用

官方规范 ↗
name
tartarusai-cli-coding-agent
description
Use TartarusAI uncensored AI coding agent via CLI for security research, automation, and code generation without policy filters
  1. 发现技能客户端向 Agent 提供名称与描述目录。
  2. 匹配与调用用户指定或任务匹配后,载入 SKILL.md 指令。
  3. 按需加载按步骤读取参考文档、使用脚本与素材。

具体调用语法与可用工具以目标 Agent 客户端为准。 查看调用机制说明 ↗

安装这个技能

Skills CLI ↗

先选择目标 Agent 和安装范围,保留技能包的附属文件,安装后检查客户端能否发现该技能。

交给 Agent 安装

复制安装指令给支持 Agent Skills 的代理,确认其中的目标目录与客户端匹配。

把 Agent Skill「tartarusai-cli-coding-agent」安装到我的项目:SKILL.md 原文与官方 description 见 https://zicq.com/zh/skills/skl-3f4e1aaa04b0462f-%E5%A1%94%E5%A1%94%E9%B2%81%E8%B5%9B%E5%85%8B%E5%88%A9%E7%BC%96%E7%A0%81%E4%BB%A3%E7%90%86.html
请存为 .cursor/skills/tartarusai-cli-coding-agent/SKILL.md 或 .claude/skills/tartarusai-cli-coding-agent/SKILL.md,frontmatter 的 name 与 description 保持原样,不要改写。

GitHub 完整包 ↗

终端安装 · Skills CLI

需要 Node.js 与 npx。先查看仓库技能列表,确认实际名称。

npx skills add 'https://github.com/reason-machines/devtools-skills' --list

npx skills add 'https://github.com/reason-machines/devtools-skills' --skill 'tartarusai-cli-coding-agent'

CLI 会交互选择目标 Agent,默认安装到项目;用户级安装使用 -g。先通过查看命令核对仓库内容,再用 npx skills list 检查已安装技能。

阅读排版
--- name: tartarusai-cli-coding-agent description: Use TartarusAI uncensored AI coding agent via CLI for security research, automation, and code generation without policy filters triggers: - use tartarus ai to generate code - run tartarusai cli for this security research - generate code with tartarus coder - use uncensored ai coding agent - help me with tartarusai cli - run tartarus ai in terminal - use tartarusai for security tooling - generate pentest code with tartarus --- # TartarusAI CLI Skill > Skill by [ara.so](https://ara.so) — Devtools Skills collection. ## Overview **tartarusai-cli** is the official terminal client for TartarusAI, an uncensored AI coding agent designed for security research, automation, and edge-case development work. Unlike mainstream AI models, TartarusAI doesn't apply content policy filters to refuse legitimate security research, penetration testing tools, or complex automation tasks. **Key capabilities:** - 256K token context window for processing entire repositories - Uncensored responses for security research and pentest tooling - Crypto-only billing (no credit card required) - Terminal-based interface for coding workflows - Suitable for vulnerability research, incident response automation, and lab environments **Not suitable for:** - Malware/exploit development for unauthorized use - DRM bypass or software piracy - Attacking systems without permission ## Installation ### Linux (x86_64) ```bash # Download and install curl -L -o tartarusai-cli https://github.com/Tartarus-AI/tartarusai-cli/releases/latest/download/tartarusai-cli-linux-x64 chmod +x tartarusai-cli sudo mv tartarusai-cli /usr/local/bin/ # Verify installation tartarusai-cli --version ``` ### Windows (x86_64) ```powershell # PowerShell Invoke-WebRequest https://github.com/Tartarus-AI/tartarusai-cli/releases/latest/download/tartarusai-cli-windows-x64.exe -OutFile tartarusai-cli.exe # Add to PATH or run from current directory .\tartarusai-cli.exe ``` ### macOS and ARM Contact `[email protected]` for macOS and ARM builds. ## Configuration ### 1. Create Account and Generate Token 1. Visit [dash.tartarusai.dev/account](https://dash.tartarusai.dev/account) 2. Create account (crypto payment required for billing) 3. Generate a CLI token from dashboard ### 2. Configure Local Token Create configuration file at: - **Linux/macOS**: `~/.tartarus/cli-token.json` - **Windows**: `%USERPROFILE%\.tartarus\cli-token.json` ```json { "endpoint": "https://api.tartarusai.dev", "token": "your-cli-token-here", "user_email": "[email protected]" } ``` **Security best practice:** ```bash # Linux/macOS: Restrict file permissions chmod 600 ~/.tartarus/cli-token.json # Verify permissions ls -la ~/.tartarus/cli-token.json ``` ### Environment Variable Alternative ```bash # Set token via environment variable export TARTARUS_CLI_TOKEN="your-token-here" export TARTARUS_ENDPOINT="https://api.tartarusai.dev" # Run without config file tartarusai-cli ``` ## Usage Patterns ### Interactive Mode ```bash # Start interactive session tartarusai-cli # CLI will present an interactive prompt # Type your coding request and press enter ``` ### Common Use Cases #### 1. Security Research Tool Development ```bash # Example: Request a port scanner for authorized testing tartarusai-cli # At prompt: > Create a Python port scanner that scans ports 1-1024 on localhost > with timeout handling and service detection. Include proper error > handling and logging. ``` #### 2. Incident Response Automation ```bash # Example: Log analysis automation > Write a Bash script that parses /var/log/auth.log for failed SSH > login attempts, extracts IP addresses, and generates a summary report > with geolocation data using ip-api.com ``` #### 3. CVE Proof-of-Concept (Lab Environment) ```bash # Example: Research known vulnerability > Create a Python proof-of-concept for CVE-2021-44228 (Log4Shell) > that demonstrates the vulnerability in a controlled lab environment. > Include both the vulnerable server component and the exploit trigger. > Add clear documentation about patching. ``` #### 4. API Security Testing ```bash # Example: API fuzzing tool > Generate a Python script using requests library to test API endpoints > for common vulnerabilities: SQL injection, XSS in JSON responses, > authentication bypass. Target endpoint from environment variable > API_TEST_TARGET. ``` #### 5. Automation Script Generation ```bash # Example: Credential rotation > Create a Python script that rotates API keys across multiple services. > Read service configs from YAML, revoke old keys via REST APIs, generate > new keys, update .env files, and send Slack notification on completion. ``` ## Code Examples ### Example 1: Network Scanner (Python) Request to TartarusAI: ``` Create a Python network scanner that accepts CIDR notation, performs TCP SYN scan on common ports, identifies services, and outputs JSON. Use scapy for packet manipulation. ``` Expected output structure: ```python #!/usr/bin/env python3 import scapy.all as scapy import json import sys from ipaddress import ip_network def scan_network(target_cidr, ports=[21,22,80,443,3306,8080]): """Scan network range for open TCP ports""" results = [] for ip in ip_network(target_cidr): host_result = {"ip": str(ip), "open_ports": []} for port in ports: # SYN scan implementation pkt = scapy.IP(dst=str(ip))/scapy.TCP(dport=port,flags="S") resp = scapy.sr1(pkt, timeout=1, verbose=0) if resp and resp.haslayer(scapy.TCP): if resp[scapy.TCP].flags == 0x12: # SYN-ACK host_result["open_ports"].append(port) # Send RST to close connection scapy.send(scapy.IP(dst=str(ip))/scapy.TCP(dport=port,flags="R"), verbose=0) if host_result["open_ports"]: results.append(host_result) return results if __name__ == "__main__": target = sys.argv[1] if len(sys.argv) > 1 else "192.168.1.0/24" results = scan_network(target) print(json.dumps(results, indent=2)) ``` ### Example 2: Log Parser for Security Events (Bash) Request: ``` Write a Bash script that monitors auth.log for brute force attempts, tracks failed login counts per IP, and alerts when threshold exceeded. ``` Expected pattern: ```bash #!/bin/bash LOG_FILE="${AUTH_LOG:-/var/log/auth.log}" THRESHOLD="${ALERT_THRESHOLD:-5}" OUTPUT_FILE="/tmp/ssh_bruteforce_ips.txt" # Extract failed SSH attempts grep "Failed password" "$LOG_FILE" | \ awk '{print $(NF-3)}' | \ sort | uniq -c | sort -nr | \ awk -v threshold="$THRESHOLD" '$1 >= threshold {print $2, $1}' > "$OUTPUT_FILE" # Alert if threats found if [ -s "$OUTPUT_FILE" ]; then echo "⚠️ SSH brute force detected:" cat "$OUTPUT_FILE" # Optionally: webhook notification if [ -n "$SLACK_WEBHOOK" ]; then curl -X POST "$SLACK_WEBHOOK" \ -H 'Content-Type: application/json' \ -d "{\"text\":\"SSH brute force alert: $(wc -l < "$OUTPUT_FILE") IPs detected\"}" fi else echo "No brute force attempts detected" fi ``` ### Example 3: JWT Token Analyzer (Python) Request: ``` Create a Python script that decodes JWT tokens, checks for common vulnerabilities (none algorithm, weak secrets), and validates claims. ``` Expected structure: ```python #!/usr/bin/env python3 import jwt import json import sys from datetime import datetime def analyze_jwt(token): """Analyze JWT token for security issues""" analysis = {"vulnerabilities": [], "claims": {}} # Decode without verification to inspect try: header = jwt.get_unverified_header(token) payload = jwt.decode(token, options={"verify_signature": False}) analysis["header"] = header analysis["claims"] = payload # Check algorithm if header.get("alg", "").lower() == "none": analysis["vulnerabilities"].append("CRITICAL: 'none' algorithm allows signature bypass") # Check expiration if "exp" in payload: exp_time = datetime.fromtimestamp(payload["exp"]) if exp_time < datetime.now(): analysis["vulnerabilities"].append("Token expired") else: analysis["vulnerabilities"].append("No expiration claim (exp) found") # Check for weak claims if "admin" in payload or "role" in payload: analysis["vulnerabilities"].append("WARNING: Privilege claims in token (check for tampering)") return analysis except jwt.DecodeError as e: return {"error": f"Invalid JWT: {str(e)}"} if __name__ == "__main__": if len(sys.argv) < 2: print("Usage: ./jwt_analyzer.py ") sys.exit(1) result = analyze_jwt(sys.argv[1]) print(json.dumps(result, indent=2, default=str)) ``` ## Best Practices ### 1. Context Management TartarusAI supports 256K context — leverage this for whole-repository analysis: ```bash # Prepare context file with repository structure find . -type f -name "*.py" | head -20 | xargs cat > /tmp/repo_context.txt # In prompt, reference the context > Analyze the Python codebase provided. Identify potential security issues > in authentication handling and suggest refactoring. Focus on SQL injection > and authentication bypass vulnerabilities. ``` ### 2. Iterative Development ```bash # Initial request > Create a basic HTTP server in Python that serves static files # Follow-up refinement > Add authentication middleware using JWT tokens from environment variable > JWT_SECRET. Include rate limiting (10 req/min per IP) using Redis. # Further iteration > Add logging of all authentication attempts to syslog with structured > JSON format including timestamp, IP, username, and success/failure. ``` ### 3. Environment-Aware Code Generation Always request code that uses environment variables for sensitive data: ```bash # Good request > Create PostgreSQL backup script that uses $DB_HOST, $DB_USER, > $DB_PASSWORD from environment. Backup to S3 using $AWS_ACCESS_KEY_ID > and $AWS_SECRET_ACCESS_KEY. # Avoid > Create PostgreSQL backup script with username 'admin' password 'password123' ``` ### 4. Testing and Validation Request test cases alongside implementation: ```bash > Generate a Python port scanner with scapy. Include pytest unit tests > that mock network responses and validate timeout handling, error cases, > and result parsing. Use pytest-mock for scapy.sr1 mocking. ``` ## Troubleshooting ### Token Authentication Issues **Problem**: `Authentication failed` or `Invalid token` ```bash # Verify token file exists and has correct structure cat ~/.tartarus/cli-token.json # Check file permissions (should be readable by user only) ls -la ~/.tartarus/cli-token.json # Test with explicit token path TARTARUS_CONFIG=~/.tartarus/cli-token.json tartarusai-cli ``` **Problem**: `Token expired` or billing issues - Visit [dash.tartarusai.dev/account](https://dash.tartarusai.dev/account) to check account balance - Crypto payment confirmation typically takes ~30 seconds - Generate new CLI token if current token was revoked ### Network Connectivity **Problem**: `Cannot connect to endpoint` ```bash # Test API connectivity curl -v https://api.tartarusai.dev/health # Check proxy settings if behind corporate firewall export HTTP_PROXY="http://proxy.corp.com:8080" export HTTPS_PROXY="http://proxy.corp.com:8080" tartarusai-cli ``` ### Response Quality Issues **Problem**: Responses are too generic or refuse task 1. **Be specific about use case**: "for authorized penetration testing in lab environment" 2. **Provide context**: Include that you own/have permission for the target system 3. **Reference legitimate frameworks**: "using Metasploit-style approach" or "similar to OWASP ZAP functionality" **Example of well-formed prompt:** ``` I'm conducting authorized penetration testing on my company's staging environment (written permission obtained). Create a Python script that tests our web application at $TARGET_URL for OWASP Top 10 vulnerabilities. Include SQL injection, XSS, and authentication bypass tests. Output results in JSON format compatible with DefectDojo import. ``` ### Large Context Handling **Problem**: Response truncated or context exceeded ```bash # Split large requests into phases # Phase 1: Architecture > Design the architecture for a log aggregation system that processes > 10k events/sec. Use Python, Redis, and PostgreSQL. Output component > diagram and data flow. # Phase 2: Implementation (reference phase 1) > Based on the architecture designed, implement the Redis event consumer > component. Include connection pooling, retry logic, and graceful shutdown. # Phase 3: Testing > Generate integration tests for the Redis consumer using pytest-asyncio > and fakeredis. Mock PostgreSQL with pytest-postgresql. ``` ## Integration with Development Workflow ### Git Commit Message Generation ```bash # Generate commit message from diff git diff > /tmp/changes.diff # In tartarusai-cli: > Analyze this git diff and generate a conventional commit message. > Include type (feat/fix/refactor), scope, and detailed body explaining > the changes and their impact. > [paste diff content] ``` ### Code Review Automation ```bash # Request security-focused code review > Review this Python Flask API code for security vulnerabilities. > Check for: SQL injection, authentication bypass, CSRF, insecure > deserialization, hardcoded secrets, and improper error handling. > Provide specific line numbers and remediation code. > [paste code] ``` ### Documentation Generation ```bash > Generate comprehensive README.md for this Python security scanner. > Include: overview, installation, usage examples, command-line arguments, > environment variables, legal disclaimer for authorized use only, and > example output. Follow security tool documentation best practices. ``` ## Additional Resources - **Website**: [tartarusai.dev](https://tartarusai.dev) - **Dashboard**: [dash.tartarusai.dev](https://dash.tartarusai.dev) - **Discord**: [discord.gg/GfzePawBBd](https://discord.gg/GfzePawBBd) - **GitHub**: [github.com/Tartarus-AI/tartarusai-cli](https://github.com/Tartarus-AI/tartarusai-cli) - **Support**: [email protected] ## Ethical Use Guidelines TartarusAI is designed for legitimate security research and development. Always ensure: 1. ✅ You have **written authorization** to test target systems 2. ✅ Testing is performed in **isolated lab environments** when learning 3. ✅ Research focuses on **defensive security** and patching 4. ✅ Code is used for **authorized penetration testing** or **your own systems** 5. ❌ Never use for unauthorized access, malware distribution, or illegal activities The tool generates code for educational and professional security work. Use responsibly and in compliance with applicable laws and regulations.

相关技能

安全测试

Skill Vetter

人工智能特工的安全第一技能审查。 在安装来自 ClawdHub, GitHub 或其他来源的任何技能之前使用 。 检查红旗,许可范围,可疑图案.

安全测试

Moltguard

MoltGuard——由OpenGuardrails担任OpenClaw保安. 安装 MoltGuard 以保护你和你的人类免受迅速注射、数据过滤和恶意命令的影响。 资料来源:https://github.com/openguardrails/openguardrails/tre…

安全测试

Security Auditor

用于审查安全漏洞代码,实施认证流,审计OWASP Top 10,配置CORS/CSP头,处理秘密,输入验证,SQL注射预防,XSS保护,或任何与安全相关的代码审查.

安全测试

Skill Vetter

安装任何AI代理技能前的安全审查协议. 红旗侦取入证盗取,疏通密码,放出. 风险分类:低/中/高/高。 编制结构化审查报告。 不先运行此功能, 绝不安装不信任的技能 .