AI Skill Hub 强烈推荐:AI安全扫描 是一款优质的Agent工作流。AI 综合评分 8.0 分,在同类工具中表现稳健。如果你正在寻找可靠的Agent工作流解决方案,这是一个值得深入了解的选择。
AI安全扫描 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
AI安全扫描 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
# 方式一:go install(推荐) go install github.com/chainreactors/aiscan@latest # 方式二:从源码编译 git clone https://github.com/chainreactors/aiscan cd aiscan go build -o aiscan . # 方式三:下载预编译二进制 # 访问 Releases 页面下载对应平台二进制文件 # https://github.com/chainreactors/aiscan/releases
# 查看帮助 aiscan --help # 基本运行 aiscan [options] <input> # 详细使用说明请查阅文档 # https://github.com/chainreactors/aiscan
# aiscan 配置说明 # 查看配置选项 aiscan --config-example > config.yml # 常见配置项 # output_dir: ./output # log_level: info # workers: 4 # 环境变量(覆盖配置文件) export AISCAN_CONFIG="/path/to/config.yml"
<p align="center"> <img src="web/assets/logo.svg" width="180" alt="cyber logo"> <h1 align="center">cyber-harness</h1> <p align="center">A general-purpose agent harness for cyber workflows — everything is an extension</p> </p>
<p align="center"> <a href="https://github.com/chainreactors/cyber-harness/releases"><img src="https://img.shields.io/github/v/release/chainreactors/cyber-harness?style=flat-square&color=00E59B" alt="Release"></a> <a href="https://github.com/chainreactors/cyber-harness/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/chainreactors/cyber-harness/ci.yml?branch=master&style=flat-square&label=CI" alt="CI"></a> <a href="https://github.com/chainreactors/cyber-harness/releases"><img src="https://img.shields.io/github/downloads/chainreactors/cyber-harness/total?style=flat-square&color=00B4D8" alt="Downloads"></a> <a href="https://github.com/chainreactors/cyber-harness/blob/master/LICENSE"><img src="https://img.shields.io/badge/license-AGPL--3.0-blue?style=flat-square" alt="AGPL-3.0"></a> <a href="https://github.com/chainreactors/cyber-harness/stargazers"><img src="https://img.shields.io/github/stars/chainreactors/cyber-harness?style=flat-square&color=yellow" alt="Stars"></a> </p>
<p align="center"> <a href="README_CN.md">中文文档</a> </p>
---
cyber-harness is a general-purpose agent harness for cyber workflows. Model reasoning, tool execution, knowledge, observation, and collaboration are assembled through the same extension lifecycle. Tasks and installed extensions determine the tools and workflows to use. Use the reference distribution or compose the capabilities your own Go application needs.
Choose an entry point for your task:
| Entry point | Use it for |
|---|---|
aiscan / aiscan-full | Agent tasks, scanners, proxy routing and IOA; the full edition adds Web, browser automation, passive recon and crawling |
[cyber-audit](cmd/audit/README.md) | Model-led source-code and binary audits, evidence collection and reports |
cyber-web (from source) | A profile-neutral Web Hub for sessions, events and artifacts from scan, audit or custom AOP nodes |
agent | A minimal local Agent built from source, with files, terminal, Skills and local subagents |
Use only on explicitly authorized targets.
Run these commands from the repository root. Outputs go to bin/; Windows binaries have an .exe suffix.
| Command | Output | Capabilities | Frontend required |
|---|---|---|---|
make / make standard | bin/aiscan | Agent, core scanners, proxy, Skills and IOA | No |
make full / make web-build | bin/aiscan-full | Standard capabilities plus Web, browser, passive recon and crawling | Yes |
make audit | bin/cyber-audit | Source-code and binary audit; built from the independent cmd/audit module | No |
make agent | bin/agent | Local Agent, files, terminal, Skills and local subagents | No |
make all | bin/aiscan and bin/aiscan-full | Standard and full editions | Yes |
For targets that require the frontend, install Node.js/npm (CI uses Node.js 22), then install dependencies once before building:
npm --prefix web/frontend ci
make
make full
These targets use CGO_ENABLED=0. make web-build builds the full edition; make web also starts its Web service. Build tags are maintained in editions.env, and the build steps in Makefile. To build the separate cyber-web Hub from source, run make frontend, then CGO_ENABLED=0 go build -tags "full sqlite" -o bin/cyber-web ./cmd/cyber-web. Native recording uses make record, requires CGO and its SDK, and has separate build instructions.
make agent needs only Go and the Makefile toolchain. It excludes scanner, proxy, IOA, browser, recording and Web dependencies. After configuring a model, run it directly with -p:
make agent
./bin/agent -p "Read the current directory and explain its structure; do not modify files"
To build without Make, use PowerShell from the repository root:
$env:CGO_ENABLED = "0"
go build -trimpath -buildvcs=false -ldflags "-s -w" -o bin/agent.exe ./cmd/agent
.\bin\agent.exe -p "Read the current directory and explain its structure; do not modify files"
Resource embedding switches · External tools and offline distribution · Native recording.
aiscan agent -p "Read the current directory and explain its structure; do not modify files" -o first-run.jsonl
Run cyber-web init or aiscan init to configure your model in ~/.cyber/cyber.yaml. For project overrides, add --project. These are shared cyber-harness commands, also available in the minimal agent CLI. Existing files are preserved. Configuration guide.
Example cyber.yaml:
llm:
provider: openai
base_url: https://api.deepseek.com/v1
model: deepseek-chat
Set CYBER_API_KEY to your credential and use the endpoint and model available to your account. For Anthropic-compatible services, use provider: anthropic and the corresponding settings. See the configuration reference for protocols, profiles and precedence.
aiscan 是一款 Agentic Security Scanner(智能体安全扫描器),它将 AI 驱动的侦察能力与确定性的扫描技术完美结合。通过引入 AI Agent 的思维模式,aiscan 不仅能执行传统的安全扫描,还能像安全专家一样进行智能化的资产探测与漏洞分析,为开发者和安全从业者提供更具洞察力的自动化安全评估能力。请注意,当前版本仍处于预览阶段,API 和功能可能会随版本迭代而变化。
aiscan 集成了 AI 驱动的智能侦察与传统确定性扫描工具,能够实现从资产发现到漏洞验证的全流程自动化。它支持多阶段的自动化链式任务,既能利用传统工具进行快速扫描,也能通过 LLM 进行结果的智能验证、公开 CVE 关联分析以及动态测试,从而在保证扫描效率的同时,大幅提升漏洞发现的准确性与深度。
您可以通过源码编译的方式安装 aiscan。首先使用 git 克隆仓库:`git clone https://github.com/chainreactors/aiscan.git` 并进入目录。随后使用 Go 进行构建:执行标准构建命令 `go build -o aiscan ./cmd/aiscan` 即可获得基础版本;若需包含 Playwright、Katana 及被动扫描功能的完整版,请使用 `go build -tags full -o aiscan-full ./cmd/aiscan` 进行编译。
aiscan 支持多种运行模式以适应不同的扫描场景。您可以根据需求选择快速模式(Quick mode)进行快速的暴露面映射,或者使用全量模式(Full mode)进行深度爬取与扩展覆盖。通过命令行工具,您可以灵活启动 Agent 任务,实现从端口发现到 Web 探测,再到弱口令检测与 POC 验证的自动化安全流水线。
aiscan 的 LLM 配置非常灵活,支持通过环境变量或配置文件进行管理。您可以通过设置 `OPENAI_API_KEY` 等环境变量来快速配置。此外,您也可以在 `aiscan.yaml` 配置文件中定义 `llm` 模块,指定 `provider`(如 openai)、`api_key` 以及使用的 `model`(如 gpt-4o),以便程序调用不同的 AI 模型进行智能分析。
aiscan 提供了强大的 CLI 命令行接口,允许用户通过参数直接指定 AI 服务提供商。例如,在执行 `aiscan agent` 命令时,可以通过 `--provider` 指定服务商(如 deepseek),利用 `--base-url` 配置 API 地址,并通过 `--api-key` 和 `--model` 参数完成身份验证与模型选择,实现高度自定义的 AI 驱动扫描任务。
aiscan 的核心工作流采用多阶段自动链式设计(Multi-stage auto-chaining),涵盖了从端口发现(port discovery)到 Web 探测(web probing)、弱口令检测(weak credentials)以及 POC 验证(POC detection)的完整流程。该流水线在基础阶段无需 LLM 即可运行,确保了扫描的高效性;在进阶阶段,则可选配 AI 驱动的结果验证、公开 CVE 关联分析及动态测试,实现确定性扫描与智能分析的深度融合。
高质量的AI安全扫描工具
该工具使用 AGPL-3.0 协议,商用场景请仔细阅读协议条款,必要时咨询法律意见。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
⚠️ AGPL 3.0 — 最严格的 Copyleft,网络服务端使用也需开源,SaaS 使用受限。
总体来看,AI安全扫描 是一款质量优秀的Agent工作流,在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态,建议收藏备用,结合自身场景选择合适时机引入使用。
| 原始名称 | aiscan |
| 原始描述 | 开源AI工作流:pi-like agentic security scanner for redteam — single binary for AI-driven pente。⭐68 · Go |
| Topics | aipentestingredteamgo |
| GitHub | https://github.com/chainreactors/aiscan |
| License | AGPL-3.0 |
| 语言 | Go |
收录时间:2026-06-24 · 更新时间:2026-06-26 · License:AGPL-3.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端