Search tool to find email addresses by abusing search

emails, leads, search, engine, crawler, intel, automation, harvester, osint, penetration-testing
pip install frisbee==0.0.7



Frisbee is a small utility to collect email addresses from search engines and other free-form text sources. Frisbee makes it simple to find email addresses posted on the web by taking user-fed input and translating it into an automated search query. Users can extend frisbee by adding modules for new search engines or other obscure data sources.

Quick Start

Install the library:

pip install frisbee or python install

Run a search

frisbee search -e bing -d -l 50 --greedy --save

Search in bulk

frisbee search -e bing -f domains -l 50 --save

Sample Code

This sample code shows some of the range of functionality within the module:

from frisbee import Frisbee

# Create an instance
frisbee = Frisbee(save=True)

# Describe your job
jobs = [{'engine': 'bing', 'modifier': '',
         'domain': '', 'limit': 50}]

# Execute the jobs

# Get the results
results = frisbee.get_results()

Example Output

Below is an example job result:

    "engine": "bing",
    "modifier": "",
    "domain": "",
    "limit": 50,
    "results": {
        "start_time": "2018-12-13 16:54:15",
        "end_time": "2018-12-13 16:54:19",
        "emails": [
        "duration": "4",
        "processed": 44
    "project": "zealous_kirch"


  • Ability to search for email addresses from search engine results
  • Modular design that can be extended easily to include new sources
  • Modifier options that can filter or target search query
  • Limit option to reduce the number of results parsed
  • Greedy option to learn from collected results and fuzzy to find related
  • Save output describing job request and results
  • Individual or bulk look-ups using the command line utility



  • Feature: Added a bulk option to the command line tool to ease usage
  • Change: Replaced multiprocessing with concurrent.futures to simplify logic
  • Change: Split logic of dynamic module loading and future work outside of the Frisbee class
  • Change: Reverted back to the BS4 parsing versus raw text
  • Change: Replaced the regular expression processing to be more efficient
  • Change: Progressively save results as they come in to avoid any losses from a deadlock
  • Change: Randomize the top-level directory to avoid conflicts


  • Feature: Clean SERPs to remove files or other formats we can't inspect
  • Change: Use text extraction instead of BS4 HTML parsing to get body of websites (ensures clean email extraction)
  • Change: Increased logging and timeout parameters


  • Feature: Added typing to the core code
  • Feature: Added a fuzzy flag to find related domains


  • Feature: Activated greedy option to save and output to screen
  • Bugfix: Wrapped loading of HTML for cases where data is dirty


  • Initial push!