https://github.com/striter-no/teinvcaptcha
TeinvCAPTCHA project is designed to analyze ad and fishing userbots group members, dead accounts, and inactive members for further administration.
https://github.com/striter-no/teinvcaptcha
invisible-captcha moderation-bot pyrogram python telegram
Last synced: about 2 months ago
JSON representation
TeinvCAPTCHA project is designed to analyze ad and fishing userbots group members, dead accounts, and inactive members for further administration.
- Host: GitHub
- URL: https://github.com/striter-no/teinvcaptcha
- Owner: striter-no
- Created: 2025-01-25T15:31:04.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-01-26T15:19:22.000Z (4 months ago)
- Last Synced: 2025-02-11T04:13:38.067Z (3 months ago)
- Topics: invisible-captcha, moderation-bot, pyrogram, python, telegram
- Language: Python
- Homepage:
- Size: 53.8 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# TeinvCAPTCHA
## What is it
This project is designed to analyze ad and fishing userbots group members to identify bots, dead accounts, and inactive members for further administration.
## How to use it
Before you start working with the software part you need to register the application in Telegram (via [this link](https://my.telegram.org/auth?to=apps)). You need to create an application and copy api_id and api_hash from there.
Next, move `tg-config.json` from the `templates` folder to the main folder.
Paste the api_id and api_hash into the appropriate fields:
Also you need your personal Telegram UID for it.
```
{
"api_id": "1234",
"api_hash": "abcdef1234",
"your_id": 1234
}
```
To start using it, it is advisable to create `venv` and use it:
```shell
python -m venv venv
source ./venv/bin/activate # For linux
```
Next, install all the necessary libraries:
```shell
pip install -r ./reqs.txt
```
Finally, you can run the main file:
```shell
python ./main.py
```
At the beginning you will be prompted to log in to your account, log in. Pass 2FA if you have it enabled.
After in telegram you can write to __any__ chat room
```plain
!anl --group_type=TYPE --chat_id=ID [--send_back]
```
Where instead of ID is the ID of the group or supergroup
Instead of TYPE - type of group:
- `group` : Regular group
- `supergroup` : Supergroup
>TIP: You can type arguments in any order and not only in 1 line:
```
!anl
--group_type=TYPE
--chat_id=ID
[--send_back]
```
You can use `--send_back` to send the report archive back to you saved messages. For example:
```plain
!anl
--group_type group
--chat_id 1234
--send_back
```
## Where to see the ID of a group or supergroup
To view the ID you need:
1. Go to Telegram settings
2. Go to the "Advanced" tab
3. At the very bottom, click the "Experimental" tab
4. Enable the display of "Peer IDs"
## How to distinguish a group from a supergroup
- Group:
- Number of members up to 200
- No invitation link
- Private only
- Supergroup:
- Number of participants up to 200'00
- Have invitation link
- Can be either public or private
## How it works
The principle of operation for the Beta 2 version period is as follows:
1. The program gets access to an open chat room (group or supergroup)
2. The program gets all users of the group, performing initial analysis
---
Primary (plain) analysis is obtaining public information about a user, i.e.:
- Full Name
- Username
- Phone number (if available)
- When was last online
- When joined the group
- How many posts have been written
- Small version of avatar
- Presence of premium on the account
- Unique Telegram User Identification Number (Telegram UID)
- Is this a bot or not
---
3. Depending on the number of messages:
- If there are 0 or 1 (only message about joining the group or basically none), then __level of trust__ becomes 0. This is probably a bot or a dead account
- If there are up to a treshhold (50 by default), then each message is pulled out for further analysis:
- Text
- Date of change
- ID
- The date the message was written
In this case, the confidence level is 0.5
- If there are more messages than treshhold, then the trust level is maximized.
4. Finally the report made is written to a *.json file
5. Then all user avatars are downloaded by file_id for further analysis