april 9, 2023
abandonware - sorry - busy elsewhere. will recode from scratch in a xplatform language in exchange for tons of money.

october 6, 2021
this may soon no longer be abandonware. as for now it is very unstable. i'd like to recode it from scratch.

description
chatter is a proof of concept osint monitoring telegram bot for windows (server, ideally) that monitors tweet content, reddit submission titles and 4chan post content for specific keywords - as well as phrases in quotation marks. it feeds content that is discovered to your telegram group in near real-time depending on your configuration. this is an early beta release with limited features.

ping kevin@envadr.io w/ questions

how to run chatter
download and unzip https://github.com/visualbasic6/chatter/archive/master.zip.

1. create your own bot for chatter on telegram with "botfather"
   
2. set your bot's api key in \bin\config\apikey.txt - make sure to put the text bot in front of the api key telegram gives you, e.g. bot111093451:GDGks4kisDFJFlulzjser-dfsSJDFpWsd
   
3. add your bot to your group and give it admin privileges
   
4. pull your telegram group's chat id by adding the bot "RawDataBot" to your group. kick it after you've received your chat id. your chat id will be displayed in RawDataBot's message like this:
   

"chat": {
            "id": -100156734232,

where -100156734232 is your telegram group's chat id.
5. set the chat id in \bin\config\chatid.txt
6. populate \bin\config\targets.txt and \bin\config\keywords.txt accordingly. don't worry - examples are already provided in these text files.
7. run the .exe

video (click to watch)

telegram bot reporting

you may need to run cmd.exe as admin and "regsvr32 {file.ocx}" if any launch errors unrelated to absent config files appear.

the files in \config\ are self-explanatory and come with examples.

the format of targets.txt is service:area:apicreds(unnecessary):minutesbetweencrawl e.g. reddit:cryptocurrency:n/a:5

populate these files prior to running the .exe or chatter will fail. it is best to use chatter with a windows server to set it and forget it. chatter was written in classic vb with an ssl wrapper making external api requests. this is rare in 2020. because of this av may display false positives. you are free to compile from source or run chatter in an enclave - which should be standard practice for any software containing code you are unable to personally review and compile. it is recommended to abstain from broadcasting to telegram several times while populating the db with content to ignore. this is particularly useful for twitter. eventually only recent material is intercepted - though it may take awhile for old content associated with your keywords and phrases to be databased and ignored.

compile from source
vb6

roadmap, misc. notes

1. monitoring more public facing areas (google, google news, fb, youtube, etc.)
2. monitoring topic-centric private areas (telegram, discord, slack, fb groups, etc.)
3. automated this with conversational ai (yeah probably won't do this)
4. web front
5. admin cmds in telegram groups e.g. twitter black/whitelisting to reduce noise

deeper dive, local storage, immediate thoughts may 17 2020

note: chatter needs to be updated to store full urls somewhere that are associated with their keywords and timestamps - and probably also crawl the urls upon their discovery to fully mirror content. this could/should come in a nearby update.

chatter is presently best reserved for something like casually sitting in a telegram group and running collaborative intel; e.g. investors overseeing news related to specific markets. it does not store post content locally. currently you would need to make edits to chatter and perform additional crawling.

the benefit of reporting discoveries to telegram groups is that telegram crawls the urls chatter presents it with and displays relevant text and image previews associated with the content.

the 3 queries chatter makes every N minutes are

https://www.reddit.com/search.compact?q=subreddit%3A{subreddit}+{query}&sort=new&t=all
https://find.4chan.org/?q={query}&b={area, e.g. biz, b, etc.}
https://api.twitter.com/2/search/adaptive.json?&q={query}

then it stores unique url identifiers as .txt files as \db\{platform}\{identifier}.txt as to ignore them as duplicates in the future.

examples
\db\reddit\1xy9k6.txt
\db\4chan\75749943.txt
\db\twitter\692151844227911680.txt

with twitter you would be able to take something like \db\twitter\692151844227911680.txt, generate the url via https://twitter.com/{anything}/status/692151844227911680 and then scrape and db the content of the tweet with another bot. alterations to chatter would be required to associate the tweet with the keyword used to locate it.

greetz
justinakapaste.com and all of my elderly aol hackers stang for the keks and lulz