GSIL(GitHub Sensitive Information Leakage)

Monitor Github sensitive information leaks in near real time and send alert notifications.

Installation

Python3(Python2 is not tested)

$ git clone https://github.com/FeeiCN/gsil.git
$ cd gsil/
$ pip install -r requirements.txt

Configuration

gsil/config.gsil(Rename by config.gsil.example): Alarm mailbox and Github configuration

[mail]
host : smtp.exmail.qq.com
# SMTP port (Not SSL port, but will use TLS encryption)
port : 25
# Multiple senders are separated by comma (,)
mails : gsil@domain.com
from : GSIL
password : your_password
# Multiple recipients are separated by comma (,)
to : feei@feei.cn

[github]
# Whether the scanned data will be cloned to the local area immediately
# Clone to ~/.gsil/codes/ directory
clone: false

# Github Token, multiple tokens are separated by comma (,)
# https://github.com/settings/tokens
tokens : your_token

gsil/rules.gsil(Rename by rules.gsil.example): scanning rules

Generally, The best rule is the characteristic code of the intranet(Example: mogujie's extranet is mogujie.com, intranet is mogujie.org. At this time, mogujie.org can be used as a rule)

There are other similar code head characteristic code, external mailbox characteristic code, and so on

field	meaning	optional	default	describe
keyword	key word	required	-	When multiple keywords are used, space segmentation is used(Example: `'username password'`), When you need a precise search, use double(Example: `"quotesele.me"`)
ext	file suffix	optional	all suffixes	Multiple suffixes are separated by comma(Example: `java,php,python`)
mode	matching mode	optional	normal-match	`normal-match`(The line that contains the keyword is matched, and the line near the line is matched) / `only-match`(Only the lines that match the key words) / `full-match`(Not recommended for use)(The search results show the entire file)

{
    # usually using the company name, used as the first parameter to open the scan(Example:`python gsil.py test`)
    "test": {
        # General use of product name
        "mogujie": {
            # Internal domain name of the company
            "\"mogujie.org\"": {
                # mode/ext options no need to configure by default
                "mode": "normal-match",
                "ext": "php,java,python,go,js,properties"
            },
            # Company code's characteristic code
            "copyright meili inc": {},
            # Internal host domain name
            "yewu1.db.mogujie.host": {},
            # External mailbox
            "mail.mogujie.com": {}
        },
        "meilishuo": {
            "meilishuo.org": {},
            "meilishuo.io": {}
        }
    }
}

Usage

$ python gsil.py test

# Verify tokens validity
$ python gsil.py --verify-tokens

GSIL changed from the original crontab startup to a resident process, how many tokens will start how many processes. It is recommended to use supervisor to manage it.

If you are just testing, you can use the following command:

python gsil.py test

Once the scan report will not repeat the report, the cache records in ~/.gsil/ directory
There are scan results in json format in the ~/.gsil/ directory, which can be consumed by message middleware

Reference

GSIL详细介绍