Airecon：自主智能体托管大语言模型测试云端依赖

AIRecon 是 an 自主智能体 cybersecurity agent combinesself-托管 大语言模型 (Ollama)与 a Kali Linux Docker sandbox、a Textual TUI. It 是 designed 到 automate security assessments, penetration 测试,、bug bounty reconnaissance — without any API keys、云端 依赖.GitHub 779 ⭐ / 127 Fork，主要使用 Python 开发。标签：ai-agents,automation,bugbo……

项目介绍

AIRecon 是 an 自主智能体 cybersecurity agent combinesself-托管 大语言模型 (Ollama)与 a Kali Linux Docker sandbox、a Textual TUI. It 是 designed 到 automate security assessments, penetration 测试,、bug bounty reconnaissance — without any API keys、云端 依赖.

核心功能

- Privacy First — Target intelligence, tool output, and reports never leave your machine. - Caido Native — 5 built-in tools: list, replay, automate (`§FUZZ§`), findings, scope. - Full Stack — Kali sandbox + browser automation + custom fuzzer + Schemathesis API fuzzing + Semgrep SAST. - Skills Knowledge Base — 57 built-in skill files, 289 keyword → skill auto-mappings. Extended by [airecon-skills](https://github.com/pikpikcu/airecon-skills) — a community skill library with 57 additional CLI-based playbooks for CTF, bug bounty, and pentesting. - Local Security Knowledge Base — Optional [airecon-dataset](https://github.com/pikpikcu/airecon-dataset) indexes ~1.09M security records (CVEs, red team techniques, CTF writeups, nuclei templates, bug bounty payloads) into local SQLite FTS5. The LLM calls `dataset_search` autonomously before attempting unfamiliar techniques — grounding its decisions in real indexed data.

使用指南

- SQLite 内存 DB 在`~/.airecon/内存/airecon.db` storing sessions, findings, patterns, target intel, 工具 usage, model performance, skill usage,、attack-chain discoveries. - Adaptive learning 状态在`~/.airecon/learning/global_learning.json` (工具 performance stats, strategy patterns, observation log, distilled insights). - Per-target 内存 files under `~/.airecon/内存/by_target/`当 persisted, containing endpoints, vulns, WAF bypasses, sensitive params,、auth endpoints. - Payload 内存 snapshots cansaved under `workspace/<target>/payload_memory.json`当会话 persistence runs. How 它 affects behavior: - On 会话 start, 内存 上下文是 injected (target intel, similar findings, learned patterns, 工具 reliability). - Every 8 iterations, learned patterns、similar findings canre-injected 基于 detected tech. - Adaptive 工具 ranking uses historical success/failure 到 order tools、suggest strategies. - Payload 内存 (when enabled) skips payloads repeatedly failed 为 the same target/param.

项目信息

项目地址

GitHub 仓库：<https://github.com/pikpikcu/airecon>

> 完整 URL：[https://github.com/pikpikcu/airecon](https://github.com/pikpikcu/airecon)

如需查看完整源码、安装命令、使用示例、API 文档，请直接访问上方项目地址。所有信息均以 GitHub 仓库为准。