Core Features

Technical overview of BotServer capabilities. For the complete feature matrix, see Feature Reference.

Multi-Channel Communication

All channels share the same conversation logic through a unified abstraction.

Authentication & Sessions

• Password Hashing: Argon2 with secure defaults
• Session Tokens: Cryptographically secure generation
• Session Persistence: Survives restarts (database-backed)
• User Isolation: Each user has isolated session state

BASIC Scripting

Built on the Rhai engine with custom keywords:

TALK "Hello!"                    ' Output
HEAR name AS NAME                ' Input with validation
result = LLM "Summarize: " + text  ' AI integration
USE KB "docs"                    ' Knowledge base

Scripts stored as .gbdialog files in bot packages.

LLM Integration

Features: prompt templates, context injection, token management, cost optimization.

Knowledge Base

• Vector Database: Qdrant for semantic search
• Document Processing: PDF, DOCX, HTML, TXT extraction
• Auto-Indexing: Documents indexed on upload
• Context Retrieval: Automatic injection into LLM prompts

Storage

Object Storage (S3-Compatible)

• Bucket management
• Secure credential-based access
• Template and asset storage

File Monitoring

• Real-time change detection
• Automatic processing triggers
• Event-driven workflows

Database

PostgreSQL with Diesel ORM:

• Connection pooling (R2D2)
• Automatic migrations
• ACID transactions

Key tables: users, bots, sessions, messages, conversations

Automation

SET SCHEDULE "0 9 * * *"  ' Daily at 9 AM
SEND MAIL "team@company.com", "Daily Report", report

• Cron scheduling
• Event triggers
• Background jobs

Security

Optional Components

Extensibility

• Custom Keywords: Add BASIC keywords in Rust
• Tool Integration: Call external APIs from scripts
• Custom Channels: Implement new communication channels
• LLM Providers: Add new AI providers

See Also

• Feature Reference - Complete feature matrix
• AI and LLM - AI integration details
• Automation - Scheduling and triggers