PyAutoRaid

Full-stack automation for Raid: Shadow Legends — from daily task farming to UNM Clan Boss optimization with real-time battle telemetry.
Features · Architecture · Report Bug

About

PyAutoRaid started as a simple PyAutoGUI screen clicker and evolved into a full game-data automation platform. It now uses a BepInEx IL2CPP mod injected into the Unity game process to read and control everything via an HTTP API — no screen scraping needed for most tasks.

Runs headless on a Windows 10 VM (QEMU/KVM), fully automated on a cron schedule.

Built With

(back to top)

Features

Game Injection — BepInEx Mod

• HTTP API on port 6790 with 30+ endpoints
• Full hero roster (500+ heroes with stats, skills, masteries, artifacts)
• All artifacts (2,600+ including vault) with substats, glyphs, Divine enhancement
• Account data (Great Hall, Arena league, Clan level)
• Game's own stat calculator integration for exact values
• UI control: button clicks, navigation, dialog dismiss, MVVM context calls
• Artifact equip/unequip/swap/bulk-equip
• Mastery open/reset

Battle Telemetry

• Per-turn snapshots of ALL heroes (HP, TM, skills, status flags)
• Direct buff/debuff tracking — every active effect with StatusEffectTypeId, turns remaining, and who placed it
• Complete StatusEffectTypeId enum (110 effects mapped from IL2CPP dump: Unkillable=320, Block Damage=60, Poison=80, HP Burn=470, etc.)
• Stat-mod effects with exact values (DEF Down, ATK Down, Weaken)
• Absorbed damage by effect kind

Daily Automation (11 tasks)

• Gem Mine, Shop Rewards, Timed Rewards, Quests, Clan, Inbox
• Arena (10 battles with weakest-opponent targeting via memory reading)
• Clan Boss (difficulty selection, instant fight detection)
• Market Shards, Artifact Auto-Sell, Dungeon Farming (20 runs)
• Three-tier control: Mod API > Memory Reading > Screen Fallback

Clan Boss Simulator

• DWJ-accurate turn-by-turn engine (tick-based TM, tie-breaking, buff timing)
• Force-Affinity damage caps calibrated against real battle data
• --use-current-gear — uses hero's actual equipped artifacts for exact stat matching
• --validate-against <log> — compares sim output vs real battle log
• Survival model: HP tracking, Gathering Fury, Lifesteal, Leech, Ally Protect
• Monte Carlo mode for RNG variance testing

CB Team Optimizer

• Evaluates 3,600+ team combinations at full potential
• Constraint-based gear assignment with speed tune targets
• Myth Eater speed tune presets (Maneater 287-290, Demytha 171-174, etc.)
• 343 heroes auto-profiled from game skill data with exact StatusEffectTypeIds

Battle Log Analyzer

• UK/BD coverage gap detection per boss turn
• Buff/debuff lifecycle with turn-by-turn transitions
• Skill rotation inference from cooldown tracking
• Damage attribution per hero per source type

Memory Reader

• Direct IL2CPP process memory reading via pymem
• Resources, battle state, current screen (497 ViewKeys), arena opponents
• No game modification needed — read-only observation

(back to top)

Architecture

                    +-------------------+
                    |   Cron Schedule   |
                    |  7am / 1pm / 7pm  |
                    +--------+----------+
                             |
                    +--------v----------+
                    | hybrid_controller |  (Python, Win10 VM)
                    |   11 daily tasks  |
                    +--+------+------+--+
                       |      |      |
              +--------+  +---+---+  +--------+
              |           |       |           |
     +--------v---+  +---v----+  +v--------+  +--------+
     | Mod API    |  | Memory |  | Screen  |  | WinRM  |
     | port 6790  |  | Reader |  | Fallback|  | Deploy |
     +--------+---+  +--------+  +---------+  +--------+
              |
     +--------v-------------------------------------------+
     |          BepInEx IL2CPP Plugin (C#)                |
     |  RaidAutomationPlugin.cs (~8K lines)               |
     |  - HTTP server, Harmony hooks, IL2CPP reflection   |
     |  - Battle telemetry, buff/debuff tracking          |
     |  - Artifact/mastery commands                       |
     +----------------------------------------------------+
              |
     +--------v---------+
     |  Raid.exe         |
     |  Unity 6000.0.60  |
     +-------------------+

     +-------------------+       +-------------------+
     |   CB Simulator    |       |  Gear Optimizer   |
     |   cb_sim.py       |<----->|  gear_optimizer.py|
     +-------------------+       +-------------------+
              |
     +--------v----------+
     | Battle Analyzer   |
     | battle_log_analyze |
     +-------------------+

(back to top)

Getting Started

Prerequisites

• Windows 10 VM (or bare metal) with Raid: Shadow Legends installed
• Python 3.12+
• BepInEx 6.0.0-be.755 (for Unity 6000.0.60f1 support)
• .NET 6.0 SDK (for building the mod)

Installation

# Clone
git clone https://github.com/SnoopLawg/PyAutoRaid.git
cd PyAutoRaid

# Install Python deps
pip install -r requirements.txt

# Build the BepInEx mod (on the VM)
cd mod/bepinex
dotnet build RaidAutomationPlugin.csproj -c Release
# Copy DLL to BepInEx/plugins/

See CLAUDE.md for full setup instructions including BepInEx configuration, port forwarding, and VM deployment.

Usage

# Daily automation (runs all 11 tasks)
python Modules/hybrid_controller.py

# Fetch fresh game data from mod API
python tools/refresh_data.py

# CB simulator with current gear
python tools/cb_sim.py --team "Maneater,Demytha,Ninja,Geomancer,Venomage" --use-current-gear

# Validate sim against real battle log
python tools/cb_sim.py --team "..." --use-current-gear --validate-against battle_log.json

# Rank all possible CB team combos
python tools/cb_potential.py

# Optimize gear for a team
python tools/gear_optimizer.py --team "..."

# Analyze a captured battle log
python tools/battle_log_analyze.py battle_log.json

(back to top)

Roadmap

• Screen automation via PyAutoGUI
• Direct IL2CPP memory reading via pymem
• BepInEx mod with HTTP API (30+ endpoints)
• Battle telemetry with direct buff/debuff tracking
• Complete StatusEffectTypeId enum from IL2CPP dump
• CB turn-by-turn simulator (DWJ-accurate)
• CB team optimizer (3,600+ combos evaluated)
• Gear optimizer with speed tune constraints
• Sim validation against real battle logs
• Relic bonus calculation
• Leader skill auras in sim
• 4-piece set bonuses in gear optimizer
• Full auto dungeon team optimizer

(back to top)

License

No License. See LICENSE.txt for more information.

Contact

Project Link: https://github.com/SnoopLawg/PyAutoRaid

Acknowledgments

• DeadwoodJedi for speed tune theory
• BepInEx for Unity IL2CPP mod injection
• Raid for not being free-to-play friendly for a busy man

(back to top)