Discord Automation & AI Reply System(Please quote)

Заказчик: AI | Опубликовано: 24.12.2025

IProject Development Requirements 1. Overall Project Overview Project Name (Working Title) Discord Bulk Automated Posting / Messaging Task System Project Objectives 1.Support multiple Discord User Tokens 2.Operate across multiple servers and channels 3.Send messages in bulk with controllable frequency and scheduling 4.Support posting messages, posts, and comments 5.Include risk-control mechanisms (rate limits, token rotation, failure skipping) 6.Provide a visual desktop management dashboard 2. Core System Modules Each section below can be developed as an independent module. Module 1: Dashboard Features Statistics display: oActive tokens oInvalid tokens oToday’s successful tasks oToday’s failed tasks Network status: oCurrent IP oRegion (HK / SG, etc.) oProxy status Technical: oData sourced from local database oAuto-refresh every X seconds (configurable) Module 2: Token Management Token Structure { group, alias, token, status (active / invalid / frozen), use_count, success_rate, last_used } Functions Add single token Bulk import (TXT / CSV) Bulk delete / reset Validate token availability Token grouping Rotation strategies (sequential / random) Auto-mark invalid tokens Auto-pause tokens reaching risk thresholds Full token action logging Module 3: Channel Management Features Add channels (name, URL, server) Auto-detect channel type (text / forum / unknown) Bind default token or auto-select Bulk import / delete channels Module 4: Content Library (Critical) Content Types Messages Posts (with title) Comments Structure { “Title”: “Post title (optional)", "content": "Supports Discord Markdown", "images": ["img1.png", "img2.png"], "tags": ["tag1", "tag2"] } Functions Create / edit / delete content Multi-image upload Tagging and preview Bulk template import Optional Spintax support {Hi|Hello} Module 5: Task System (Core) Task Types Create post Send message Send comment (reserved) Task Configuration Tokens (single / multiple) Channels (single / multiple) Content (single / multiple) Execution count Interval (seconds) Loop execution Token rotation strategy Content rotation strategy Wait time if all tokens fail Controls Create / start / pause / stop / edit / delete Module 6: Advanced Scheduling & Risk Control Monitoring-only tokens Channel activity monitoring Channel limits: oMax active threads oMax queued tasks Automatic queueing to prevent bans Module 7: Monitoring, Stats & Logs Statistics Success / failure counts (messages, posts, comments) Overall success rate Logs Real-time logs Exportable logs Clear logs Error logs must include token, channel, error type Non-Functional Requirements No fixed intervals Random delay (±20%) Per-token frequency limits Auto slow-down on failures Delivery Windows desktop application Local SQLite database Source code or extensible interface Basic usage documentation Additional Module 1: Channel Content Collection (Images + Links) Goal Collect images, links, and text from specified Discord channels into a local content library for AI use. Collection Rules Real-time or scheduled scanning Deduplication by hash Images must be downloaded locally (no CDN-only storage) Local Image Library System Directory Structure (Recommended) /images /auto_collect/YYYY-MM /product/xxx /user_upload Image Database { image_id, local_path, hash, source, source_channel, keywords[], bind_product_id, created_at } Functions Thumbnail view & preview Keyword editing Product binding Auto-use in AI replies Image Deduplication & Cleanup Hash-based deduplication Optional unused-image cleanup AI Image Recognition + Library Matching Flow User sends image → download → AI image recognition → keyword extraction → local library match → product match → reply with image + link ❗ AI must only use local assets, not internet search. Additional Module 2: AI Auto-Reply System Triggers Keyword detection Purchase intent User image upload Text Flow User message → intent detection → product match → reply with text + image + link Image Flow User image → recognition → keyword match → local asset → reply Reply Rules (Risk Control) Randomized templates Variable insertion Cooldown per user Fallback to text-only on failure Additional Module 3: Product / File Matching Product Structure { product_name, keywords[], images[], link, priority } Matching Logic Keyword similarity Priority-based selection Additional Module 4: AI Control Panel Per-token reply limits User cooldowns Emergency AI stop Per-channel AI enable/disable Technical Notes AI modules must be replaceable Vision APIs isolated All AI features must be toggleable