skipfish是一个Linux下的web安全扫描工具,支持自定义字典扫描,可学习字典规则,识别漏洞危险级别(sql注入、shell注入、xss),并且提供可视化的统计结果,题主通常与sqlmap 组合使用,对web应用进行上线前的安全扫描和检测。
  官网:http://code.google.com/p/skipfish/ ,需翻墙。
  安装
--------------------依赖安装--------------------
yum install -y openssl-devel openssl zlib-devel zlib libidn-devel libidn
--------------------安装--------------------
wget https://skipfish.googlecode.com/files/skipfish-2.10b.tgz
tar xzvf skipfish-2.10b.tgz
cd skipfish-2.10b
make
  基本参数
-A user:pass      - use specified HTTP authentication credentials
使用特定的http验证
-F host=IP        - pretend that 'host' resolves to 'IP'
-C name=val       - append a custom cookie to all requests
对所有请求添加一个自定的cookie
-H name=val       - append a custom HTTP header to all requests
对所有请求添加一个自定的http请求头
-b (i|f|p)        - use headers consistent with MSIE / Firefox / iPhone
伪装成IE/FIREFOX/IPHONE的浏览器
-N                - do not accept any new cookies
不允许新的cookies
--auth-form url   - form authentication URL
--auth-user user  - form authentication user
--auth-pass pass  - form authentication password
--auth-verify-url -  URL for in-session detection
Crawl scope options:
-d max_depth     - maximum crawl tree depth (16)大抓取深度
-c max_child     - maximum children to index per node (512)大抓取节点
-x max_desc      - maximum descendants to index per branch (8192)每个索引分支抓取后代数
-r r_limit       - max total number of requests to send (100000000)大请求数量
-p crawl%        - node and link crawl probability () 节点连接抓取几率
-q hex           - repeat probabilistic scan with given seed
-I string        - only follow URLs matching 'string' URL必须匹配字符串
-X string        - exclude URLs matching 'string' URL排除字符串
-K string        - do not fuzz parameters named 'string'
-D domain        - crawl cross-site links to another domain 跨域扫描
-B domain        - trust, but do not crawl, another domain
-Z               - do not descend into 5xx locations 5xx错误时不再抓取
-O               - do not submit any forms 不尝试提交表单
-P               - do not parse HTML, etc, to find new links 不解析HTML查找连接
Reporting options:
-o dir          - write output to specified directory (required)
-M              - log warnings about mixed content / non-SSL passwords
-E              - log all HTTP/1.0 / HTTP/1.1 caching intent mismatches
-U              - log all external URLs and e-mails seen
-Q              - completely suppress duplicate nodes in reports
-u              - be quiet, disable realtime progress stats
-v              - enable runtime logging (to stderr)
Dictionary management options:
-W wordlist     - use a specified read-write wordlist (required)
-S wordlist     - load a supplemental read-only wordlist
-L              - do not auto-learn new keywords for the site
-Y              - do not fuzz extensions in directory brute-force
-R age          - purge words hit more than 'age' scans ago
-T name=val     - add new form auto-fill rule
-G max_guess    - maximum number of keyword guesses to keep (256)
-z sigfile      - load signatures from this file
Performance settings:
-g max_conn     - max simultaneous TCP connections, global (40) 大全局TCP链接
-m host_conn    - max simultaneous connections, per target IP (10) 大链接/目标IP
-f max_fail     - max number of consecutive HTTP errors (100) 大http错误
-t req_tmout    - total request response timeout (20 s) 请求超时时间
-w rw_tmout     - individual network I/O timeout (10 s)
-i idle_tmout   - timeout on idle HTTP connections (10 s)
-s s_limit      - response size limit (400000 B) 限制大小
-e              - do not keep binary responses for reporting 不报告二进制响应
Other settings:
-l max_req      - max requests per second (0.000000)
-k duration     - stop scanning after the given duration h:m:s
--config file   - load the specified configuration file
  使用范例
  以扫描 baidu.com 为例
  /path/skipfish -d 2 -S dictionaries/minimal.wl -o baidu http://www.baidu.com
  结果
  附上刚刚执行扫描的截图,如果有相应安全漏洞,则会直观显示出来。


扫描结果页面