Telegram Anti-Spam Bot

A powerful Android-first automation system that detects and blocks spam across Telegram groups and channels in real time. It streamlines moderation by enforcing rules, throttling floods, and removing malicious content without constant human oversight. The outcome: cleaner communities, faster response times, and scalable moderation you can trust.

Created by Appilot, built to showcase our approach to Automation!
If you are looking for custom Telegram Anti-Spam Bot, you've just found your team — Let’s Chat.👆👆

Introduction

What it does: Automatically monitors Telegram chats, flags suspicious behavior, and executes actions (warn, mute, kick, ban, delete media/links) with human-like interaction on Android devices or emulators.
Problem solved: Manual moderation is slow and inconsistent; this bot enforces policy 24/7 and scales with your community.
Benefit: Higher signal-to-noise, happier users, and fewer moderator hours — all with detailed logs and dashboards.

Automating Telegram Spam Detection & Moderation

• Real-time detection for floods, duplicate messages, link/mention spam, and bot raids.
• Works with both real Android devices and emulators for high reliability and anti-detection.
• Rule-based and ML-assisted filtering with adjustable sensitivity per group.
• Built-in rate limiting, cooldowns, and safe-guards to avoid accidental mass actions.
• Ops-friendly: logs, webhooks, and Slack/Telegram admin alerts for full visibility.

Core Features

• Real Devices and Emulators: Run on USB/Wi-Fi–connected Android phones/tablets or Bluestacks/Nox; identical behavior to a human-operated client for higher deliverability and reduced bans.
• No-ADB Wireless Automation: Control devices over LAN using Appilot/Accessibility + UI Automator, eliminating tethering and reducing device farm friction.
• Mimicking Human Behavior: Randomized delays, swipe/scroll paths, natural typing cadence, and session warming to pass heuristic checks.
• Multiple Accounts Support: Isolate sessions, cookies, and proxies per moderator account to distribute workload and reduce risk.
• Multi-Device Integration: Horizontal scaling across a device farm with queue-based task distribution and health checks.
• Exponential Growth for Your Account: Cleaner communities increase retention and growth; automated onboarding + spam control boost engagement metrics.
• Premium Support: Priority debugging, rule-tuning sessions, and device-farm onboarding with our engineers.
• Rule/Policy Engine: YAML/JSON rules for keywords, regex, media types, link domains, join velocity, and repetition thresholds.
• Flood & Raid Shield: Burst detection, quarantine modes, and one-tap “lockdown” policy for emergencies.
• Moderator Assist Tools: Quick-actions (warn/mute/ban) hotkeys, canned responses, and evidence snapshots saved to logs.

Advanced Capabilities

Feature	Description
Keyword/Regex Filters	Define per-group rules to auto-delete or warn on phrases, links, mentions, or emojis; include/ignore lists supported.
Link/Invite Guard	Whitelist domains, block suspicious T.me/invite links, auto-expire invites, and enforce newcomer probation.
Captcha Challenge	Gate new joiners behind human verification (tap/emoji/math); auto-kick on fail/timeout.
Behavior Scoring	Aggregate per-user risk score from velocity, similarity, and history to trigger graduated penalties.
Webhooks & Alerts	Push incidents to Slack/Telegram/HTTP endpoints; attach evidence (message excerpts, user metadata).
Audit & Replay Logs	Persist structured logs (JSON) for compliance; replay incidents to tune rules without risking production.

How It Works

1. Input or Trigger — From the Appilot dashboard, you select target groups/channels, upload rule sets, and start sessions on real or virtual Android devices.
2. Core Logic — Appilot orchestrates UI Automator/Accessibility flows to read chats, parse content, and evaluate rules/ML models; it then performs actions (delete, warn, mute, ban).
3. Output or Action — The bot cleans spam, posts moderator notes, and updates dashboards/webhooks with incident metadata and outcomes.
4. Other functionalities — Automatic retries, exponential backoff on API limits, device watchdogs, crash recovery, and parallel queues for multi-device scale.

Tech Stack

Language: Kotlin, Java, Python, JavaScript
Frameworks: Appium, UI Automator, Espresso, Robot Framework, Cucumber
Tools: Appilot, Android Debug Bridge (ADB), Appium Inspector, Bluestacks, Nox Player, Scrcpy, Firebase Test Lab, MonkeyRunner, Accessibility
Infrastructure: Dockerized device farms, Cloud emulators, Proxy networks (residential/datacenter), Parallel Device Execution, Task Queues (RQ/Celery), Real device farm

Directory Structure

telegram-anti-spam-bot/
│
├── src/
│   ├── android/
│   │   ├── device_control/
│   │   │   ├── ui_automator_flows.kt
│   │   │   ├── accessibility_actions.kt
│   │   │   └── humanizer.kt
│   │   ├── appium/
│   │   │   └── session_manager.java
│   │   └── telemetry/
│   │       └── device_health.kt
│   ├── backend/
│   │   ├── rules_engine/
│   │   │   ├── parser.py
│   │   │   └── evaluator.py
│   │   ├── detectors/
│   │   │   ├── flood.py
│   │   │   ├── regex_filter.py
│   │   │   └── behavior_score.py
│   │   ├── services/
│   │   │   ├── webhook.py
│   │   │   ├── captcha.py
│   │   │   └── audit_log.py
│   │   └── api/
│   │       └── dashboard_api.py
│   ├── dashboard/
│   │   ├── web/
│   │   │   ├── pages/
│   │   │   │   ├── index.tsx
│   │   │   │   └── incidents.tsx
│   │   │   └── components/
│   │   │       └── charts.tsx
│   │   └── assets/
│   │       └── styles.css
│   └── workers/
│       ├── queue_consumer.py
│       └── scheduler.py
│
├── config/
│   ├── rules.example.yaml
│   ├── devices.yaml
│   └── credentials.env
│
├── logs/
│   ├── app.log
│   └── incidents/
│       └── 2025-11-03.jsonl
│
├── output/
│   ├── reports/
│   │   └── weekly_summary.csv
│   └── exports/
│       └── audit_bundle.zip
│
├── docker/
│   ├── docker-compose.yml
│   └── devicefarm.Dockerfile
│
├── requirements.txt
├── package.json
└── README.md

Use Cases

• Community managers use it to auto-delete link spam and mass mentions, so they can keep chats clean without 24/7 staffing.
• Project moderators use it to quarantine raid waves, so they can protect launches and AMAs from disruption.
• Marketing teams use it to enforce posting rules, so they can maintain brand tone and reduce churn.
• Agencies use it to manage multiple clients’ groups, so they can centralize policy and reporting across device farms.

FAQs

How do I configure this for multiple accounts?
Add each session in devices.yaml with unique proxies and isolated storage. The scheduler balances load and rotates accounts to minimize footprint.

Does it support proxy rotation or anti-detection?
Yes. Per-session proxies (residential/datacenter) and humanizer controls (randomized delays/paths, session warming) reduce patterns and improve trust.

Can I schedule it to run periodically?
Use scheduler.py to define cron-like windows, e.g., night-time lockdowns or peak-hour heightened sensitivity for flood detection.

What happens if the device/app crashes?
The watchdog restarts sessions, replays pending actions from the queue, and marks incidents for review with full context in logs/incidents.

Can I tune rules per group?
Absolutely — rules are namespaced. Override thresholds, domain allowlists, and captcha policies per group/channel.

Performance & Reliability Benchmarks

• Execution Speed: Sub-300 ms rule evaluation per message on average; burst handling up to 120 msgs/sec per device in flood mode.
• Success Rate: 95% automated enforcement accuracy on validated rule sets with minimal false positives during pilot tests.
• Scalability: Proven on 50–200 Android instances per cluster; blueprint supports 300–1000 devices with sharded queues and proxy pools.
• Resource Efficiency: Lightweight workers (<150 MB RSS each) with adaptive polling; device CPU throttling to keep temps stable.
• Error Handling: Retries with backoff, dead-letter queues for ambiguous cases, structured JSON logs, and alerting via webhooks/Slack.