PriestyAI

PriestyAI is an autonomous AI companion, code reasoning engine, and workspace agent for Discord built on Python, discord.py, and the Google GenAI SDK.

Unlike standard conversational bots, PriestyAI features multi-turn workspace threads with Docker sandbox execution, Discord Components V2 interactive streaming, local SearXNG metasearch, custom REST webhook extensibility, and direct GitHub App integration with verified commit signing.

---

Quickstart

Prerequisites

• Discord Bot Token (with Message Content Intent enabled in the Discord Developer Portal)
• Google Gemini API Key (from Google AI Studio)
• Docker & Docker Compose (for Container setup) or Python 3.11+ (for Local setup)

---

Option A: Docker Container

Spins up both PriestyAI and the local SearXNG metasearch service in isolated containers with zero manual environment management.

1. Clone the repository:

   git clone https://github.com/Priestytheplushie/PriestyAI.git
   cd PriestyAI

2. Configure your environment:

   cp .env.example .env

   Edit .env and paste your DISCORD_TOKEN and GEMINI_API_KEY.

3. (Optional) Setup Application Emojis: Upload all bundled custom icons directly to your Discord Bot Application:

   docker compose run --rm priestyai python scripts/setup_emojis.py

4. Start the services:

   docker compose up -d --build

5. View live logs:

   docker compose logs -f priestyai

---

Option B: Local Setup

If you prefer running the Python process directly on your host machine:

1. Clone and create a virtual environment:

   git clone https://github.com/Priestytheplushie/PriestyAI.git
   cd PriestyAI
   
   python -m venv .venv
   
   # Linux / macOS
   source .venv/bin/activate
   
   # Windows (PowerShell)
   .venv\Scripts\Activate.ps1

2. Install dependencies:

   pip install -r requirements.txt

3. Configure .env:

   cp .env.example .env

   Fill in your DISCORD_TOKEN and GEMINI_API_KEY.

4. (Optional) Setup Application Emojis: Upload all bundled custom icons and badges directly to your Discord Bot Application:

   python scripts/setup_emojis.py

5. Start local SearXNG search (Docker):

   docker compose up -d searxng

6. Start the bot:

   python bot.py

---

Key Architecture & Features

1. Autonomous Agent Mode (/agent)

• Dedicated Workspace Threads: Spawns isolated private threads with dedicated workspace directories.
• Polyglot Docker Sandbox: Safely executes Python, Node.js/TypeScript, Rust, Go, C/C++, and Bash commands in isolated containers.
• Human-in-the-Loop Planning: Two-phase execution loop with plan review, clarification questions, and interactive approval cards.
• Verified GitHub Commits: Stages code changes, applies co-author attribution, and publishes branches/PRs via GitHub's Git Data API with verified commit badges.

2. Custom Tools & Webhook Extensibility

• Zero-Code REST Integrations: Server admins and users can register public APIs (e.g. game server status, clan stats, crypto feeds) via /config setting:Custom Tools.
• Dynamic Parameter Auto-Extraction: Endpoints using {param} placeholders automatically generate dynamic schema inputs for Gemini.
• Built-in Security & SSRF Protection: Hardware-level private IP blocklists and pre-flight moderation guardrails reject adversarial injections and internal network exploits.
• Role-Based Permission Controls: Restrict custom tools to specific roles or toggle user-scoped tool execution server-wide under /config setting:Tool Permissions.

3. Intelligent Routing & Resilience

• Complexity-Based Routing: Dynamically classifies query complexity, routing lightweight requests to high-speed utility models and complex engineering tasks to deep reasoning models.
• Automated Fallback Cascades: Gracefully recovers from provider rate limits, network timeouts, and upstream outages via automated model fallback chains.
• Instant Fast-Answer Interrupt: Users can click Answer Now to abort reasoning scratchpads and receive immediate streaming responses.
• SQLite WAL Mode: Concurrent non-blocking reads and writes across background watchdogs, live web terminal auto-saves, and chat threads.

4. Rich Discord Components V2 UI

• Live Code Canvas & Playground: Generates single-file scripts and multi-file .zip archives with interactive preview modals, unified diff calculation, and live web previews.
• Interactive Knowledge Quizzes: 1-by-1 stepper views with anti-spoiler thought process gates, instant scoring diagnostics, and dynamic study guide generation.
• Thought Process Inspection: Inspectable scratchpads, tool execution traces, execution timings, and raw JSON payload viewers.

5. Privacy, Security & Local Tooling

• SearXNG Metasearch: Privacy-preserving web search and high-resolution image discovery without third-party tracking.
• Field-Level Encryption at Rest: Sensitive user facts, configuration credentials, and conversation sessions are encrypted using AES-256 (Fernet with PBKDF2 key derivation).
• Automated Safety Guardrails: Pre-flight moderation filtering with zero-tolerance policy enforcement and GDPR self-service data erasure via /data.

---

Slash Commands & Context Menus

Command	Scope	Description
/ask	Global / Guild	Ask a quick question with optional ephemeral visibility.
/chat	Global / Guild	Start or continue a multi-turn conversation with persistent channel memory and session controls.
/agent	Server Text	Launch an autonomous research or engineering agent in a private thread.
/config	Server / Channel / User	Configure prompts, custom tools, permissions, reasoning depth, and personas.
/data	Global / Guild	Self-service database browser to inspect, edit, or permanently delete stored facts.
/generate	Global / Guild	Direct generation across text, local image diffusion, or native voice messages.
/scribe	Global / Guild	Translate, localize, and adapt text with custom creative tone and dialect instructions.
/schedule	Server / User	Automate recurring AI research, daily briefs, or personal DM tasks on a timeline.
/feedback	Global / Guild	Submit bug reports, feature suggestions, or tickets to the developer.
/terms	Global / Guild	Review safety policies, service guidelines, and moderation terms.
/privacy	Global / Guild	Review data retention policies, at-rest encryption, and third-party sub-processors.

Message Context Menus

• Run as Prompt: Re-evaluates any selected message or attachment as a fresh prompt.
• Branch: Creates an exploratory branch thread preserving prior conversation context.
• Retry: Generates a new alternative version of any response with version swapper controls (◀ 1/3 ▶).
• Translate: Directly translates message content into any selected language with visibility controls.
• Edit: Modifies bot responses in-place and updates stored generation history.
• View Prompt: Inspects the input prompt, author metadata, and reasoning telemetry for a bot response.
• All Commands: Comprehensive action menu providing access to Summarize, Run as Prompt, View Prompt, Retry, Branch, Delete Bot Message, and Add to Chat Context.

---

Project Structure

PriestyAI/
├── agent/                  # Autonomous agent engine, session manager, git automation
├── assets/                 # Custom emoji assets, application icons, and badges
├── commands/               # Application slash commands and context menus
├── config/                 # Settings and environment configuration
├── core/                   # Engine, client manager, encryption, router, memory manager
├── handlers/               # Chat handler, Components V2 streaming dispatcher
├── parsers/                # Discord markdown (DFM), math, emoji, mention, timestamp parsers
├── scripts/                # Asset export and application emoji auto-setup utilities
├── tools/                  # Registry and native tools (Docker sandbox, search, GitHub, math)
├── ui/                     # Components V2 layout views, modals, thought container, quizzes
├── searxng/                # Local SearXNG configuration
├── web/                    # Live canvas playground frontend template
├── bot.py                  # Discord client entrypoint & presence loop
├── requirements.txt        # Python package dependencies
├── Dockerfile              # Container definition for PriestyAI
└── docker-compose.yml      # Multi-service container orchestration

---

Terms & Privacy (Official Hosted Instance)

The policies outlined below apply exclusively to users interacting with the official, hosted instance of PriestyAI on Discord:

• Terms of Service: Governs service availability (Zero SLA disclaimer), acceptable use policies, safety filtering, and moderation enforcement.
• Privacy Policy: Details data collection, at-rest database encryption, third-party inference sub-processors (Google Gemini, Groq, OpenRouter), and user rights (GDPR right-to-erasure via /data).

Note

Note for Self-Hosters: If you are self-hosting your own independent instance of PriestyAI from this open-source repository, you are running on your own infrastructure and API credentials. You are solely responsible for managing your own data retention, privacy compliance, security, and user policies.

---

Third-Party Licenses

This project utilizes several open-source libraries, frameworks, and assets. For a full list of third-party attributions and licenses, please see the NOTICE.md file.

---

License

This project is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).