Symbiont v1.4.0

What's New in v1.4.0

Persistent Memory (MarkdownMemoryStore)

• Markdown-backed agent memory with Facts, Procedures, and Learned Patterns sections
• Daily log append with timestamped entries and retention-based compaction
• DSL memory block for declarative configuration (store, path, retention)
• REPL :memory command for inspection and management

Webhook Verification (SignatureVerifier)

• HmacVerifier with HMAC-SHA256 and constant-time comparison
• JwtVerifier with HS256 JWT token verification
• Built-in presets for GitHub, Stripe, Slack, and Custom providers
• DSL webhook block with provider presets, secrets, path, and event filtering
• Pre-handler signature verification in HttpInputServer

HTTP Input Security Hardening

• Loopback-only default binding (127.0.0.1 instead of 0.0.0.0)
• Explicit CORS origin allow-lists (replaces boolean toggle)
• JWT EdDSA (Ed25519) validation in auth middleware
• /health endpoint exempt from authentication for load balancers
• PathPrefix route matching in HTTP input routing

Skill Scanning (ClawHavoc)

• 10 built-in security rules: pipe-to-shell, wget-pipe-to-shell, env-file-reference, soul-md-modification, memory-md-modification, eval-with-fetch, fetch-with-eval, base64-decode-exec, rm-rf-pattern, chmod-777
• Custom rule support alongside defaults

Metrics & Telemetry

• FileMetricsExporter with atomic JSON writes
• OtlpExporter for OpenTelemetry endpoints
• CompositeExporter for multi-backend fan-out
• Background MetricsCollector thread

DSL Parser Fixes

• Bare identifiers in value rule (store markdown, provider github)
• Short-form duration literals (90d, 6m, 1y)

Crate Versions

Crate	Version
symbi	1.4.0
symbi-dsl	1.4.0
symbi-runtime	1.4.0
repl-core	1.4.0

Packages

• crates.io: symbi, symbi-dsl, symbi-runtime, repl-core
• npm: symbiont-sdk-js v0.6.0
• PyPI: symbiont-sdk v0.6.0

Symbiont v1.1.0

Symbiont v1.1.0

Security Hardening v2

• Per-agent API key authentication (Argon2, file-backed store)
• Per-IP rate limiting on HTTP API
• Schema-driven sensitive parameter redaction
• File locking for secret store
• Safe sandbox defaults (no shells)

DSL Improvements

• Structured diagnostic types (DslDiagnostic)
• Humantime duration parsing

New: symbi-a2ui (Alpha)

• Lit-based admin dashboard for fleet management and compliance

Channel Adapter

• jsonwebtoken v10 migration, Teams auth fix

Full Changelog

v1.0.1...v1.1.0

Symbiont v1.0.1

Symbiont v1.0.1

Highlights

AgentPin Integration

• DiscoveryMode Resolver Dispatch: Multi-strategy agent identity resolution with automatic fallback chain (StaticDocument → WellKnown → DnsRecord)
• agentpin 0.2.0: Switched from local path dependency to crates.io release with trust bundle support

MCP Server

• Real MCP server over stdio using rmcp SDK — symbi mcp now serves a proper MCP protocol endpoint

Channel Adapters

• Bidirectional Slack, Teams, and Mattermost adapters
• Channel management REST API and declarative channel {} DSL block
• Enterprise channel governance layer

Fixes

• Docker build cache invalidation for libsymbi* artifacts
• Clippy, CI test, and runtime bug fixes
• Compilation warnings resolved, Qdrant made optional

Crate Versions

Crate	Version
symbi	1.0.1
symbi-dsl	1.0.1
symbi-runtime	1.0.2
symbi-channel-adapter	0.1.0 (new)
repl-core	1.0.1

Installation

cargo install symbi

Or with Docker:

docker pull ghcr.io/thirdkeyai/symbi:latest

Full Changelog

v1.0.0...v1.0.1

Symbiont v1.0.0

Symbiont v1.0.0

The first stable release of Symbiont — an AI-native agent framework for building autonomous, policy-aware agents with scheduling, channel adapters, cryptographic identity, and multi-language SDKs.

Highlights

Production-Ready Scheduling

• Session Isolation: Per-run AgentContext with HeartbeatContextMode control (EphemeralWithSummary, SharedPersistent, FullyEphemeral)
• Jitter Support: Random delay to prevent thundering herd in multi-job deployments
• Per-Job Concurrency Guards: Configurable max_concurrent to prevent resource exhaustion
• Dead-Letter Queue: Jobs exceeding max_retries move to DeadLetter status with manual recovery workflow
• CronMetrics Observability: Prometheus-compatible metrics (runs_total, execution_duration_seconds, in_flight_jobs, dead_letter_total)
• Health Endpoint: /api/v1/health/scheduler for ops monitoring
• AgentPin JWT: ES256 cryptographic identity verification on scheduled jobs

Security Enhancements

• New security event types: CronJobDeadLettered, AgentPinVerificationFailed
• Enhanced policy enforcement: time window restrictions, capability requirements, human approval workflows, AgentPin verification

Delivery Routing

• Six built-in output channels: Webhook, Slack, Email, Custom, Stdout, LogFile
• Pluggable DeliveryRouter trait for custom routing

Channel Adapters

• Slack (community): Bidirectional messaging with webhook verification, HMAC signing, and OAuth support
• Microsoft Teams (enterprise): Azure AD OAuth2, Bot Framework API, Adaptive Cards
• Mattermost (enterprise): REST v4 API, HMAC-SHA256 webhook verification, self-hosted support
• 12 REST endpoints for channel lifecycle management (register, start, stop, health, identity mapping, audit)
• DSL channel block for declarative channel configuration

HTTP API

• 10 RESTful schedule management endpoints with OpenAPI/Swagger documentation
• 12 channel management endpoints with OpenAPI/Swagger documentation
• Complete CRUD + pause/resume/trigger/history/next-run operations
• HTTP Input module for webhook integrations with Bearer/JWT auth and rate limiting

SDK Support

• JavaScript SDK: Full TypeScript ScheduleClient and ChannelClient with CRUD operations
• Python SDK: Async ScheduleClient and ChannelClient with full API coverage

Documentation

• Comprehensive scheduling guide in 6 languages (EN, DE, ES, JA, PT, ZH-CN)
• Full API reference, DSL guide, runtime architecture docs
• Available at docs.symbiont.dev

Reliability & Performance

• Graceful shutdown with in-flight job tracking
• Exponential backoff retry logic
• ACID-guaranteed job state transitions
• Optimized tick loop and configurable global concurrency

Breaking Changes

• CronJobDefinition has new optional fields (max_concurrent, agent_pin_jwt) — existing jobs work without changes
• HeartbeatContextMode enum added — defaults to EphemeralWithSummary (previous behavior)

Migration from v0.9.0

No breaking API changes. All v0.9.0 scheduled jobs continue to work. See CHANGELOG.md for optional enhancement steps.

Crates

Published to crates.io:

• symbi v1.0.0
• symbi-runtime v1.0.1
• symbi-dsl v1.0.0
• repl-core v1.0.0
• repl-proto v1.0.0
• repl-cli v1.0.0
• repl-lsp v1.0.0
• agentpin v0.1.0

---

Full Changelog: https://github.com/ThirdKeyAI/Symbiont/blob/main/CHANGELOG.md

v0.6.1

Fixed

• Compilation Issues: Resolved crates.io publishing compilation errors
  • Fixed SecureMessage API usage with correct field names and types
  • Added missing SystemTime import for timestamp handling
  • Fixed ModelLogger API compatibility
  • Added missing dependencies (tokio, serde) to REPL crates
  • Fixed HttpInputConfig struct with required fields
  • Resolved match arm type compatibility issues in JSON-RPC server

Dependencies

• Version Specifications: Added proper version specifications to all workspace dependencies for crates.io publishing

See v0.6.0 release notes for full changes in v0.6.X

v0.6.0

Added

🧠 Complete REPL System (New)

• Interactive Development Environment: Full REPL (Read-Eval-Print Loop) system for Symbiont DSL
  • crates/repl-core: Core REPL engine with DSL evaluation, session management, and policy enforcement
  • crates/repl-cli: Interactive CLI interface and JSON-RPC server for programmatic access
  • crates/repl-proto: JSON-RPC protocol definitions for client-server communication
  • crates/repl-lsp: Language Server Protocol implementation for IDE integration
• Agent Lifecycle Management: Create, start, stop, pause, resume, and destroy agents through REPL
• Real-time Monitoring: Execution monitoring with statistics, traces, and performance metrics
• Session Management: Snapshot and restore REPL sessions with persistent state
• Policy Integration: Built-in policy checking and capability gating for security

🛡️ AI-Driven Tool Review System (New)

• Automated Security Analysis: Complete workflow for MCP tool review and signing
  • enterprise/src/tool_review/: Tool review orchestrator and components
  • AI-powered security analysis with RAG (Retrieval-Augmented Generation)
  • Human oversight integration with streamlined review interface
  • Digital signing and verification of approved tools
• Security Assessment: Risk-based analysis with configurable severity levels
  • Vulnerability detection and impact assessment
  • Automated recommendations with confidence scoring
  • Audit trail and compliance reporting

☁️ E2B Sandbox Integration (New)

• Cloud Sandbox Support: E2B.dev integration for secure code execution
  • crates/runtime/src/sandbox/e2b.rs: E2B sandbox implementation
  • Multi-tier sandbox architecture (Docker, gVisor, Firecracker, E2B)
  • Automatic tier selection based on risk assessment
  • Remote execution capabilities with enhanced isolation

📊 Enhanced Scheduler Features

• Real Task Execution: Production-grade task processing capabilities
  • Process spawning with secure execution environments
  • Resource monitoring (CPU, memory) with 5-second intervals
  • Health checks and automatic failure detection
  • Support for ephemeral, persistent, scheduled, and event-driven execution modes
• Graceful Shutdown: Enhanced termination handling
  • 30-second graceful termination period with force termination fallback
  • Resource cleanup and metrics persistence
  • Queue cleanup and state synchronization

📋 Documentation & Architecture

• Data Directory Design: Comprehensive directory structure specification
  • data_directory_structure_design.md: Enhanced data persistence architecture
  • Unified management of agent contexts, logs, prompts, and vector database storage
  • Migration utilities and backward compatibility support
• Tool Review Documentation: Complete workflow documentation
  • docs/tool_review_workflow.md: AI-driven tool review process
  • Security analysis procedures and human oversight protocols
• REPL Guide: Comprehensive user and developer documentation
  • docs/repl-guide.md: Complete REPL usage guide
  • Interactive examples and integration patterns

🔧 Release Management

• Version Bump: Updated to 0.6.0 across all workspace crates
• Documentation Updates: Updated version references in documentation and examples

Improved

Developer Experience

• Unified Workspace: Enhanced project organization with REPL crates
  • Consistent versioning across all workspace members
  • Improved dependency management between crates
• IDE Integration: Language Server Protocol support for enhanced development
  • Syntax highlighting and completion for Symbiont DSL
  • Real-time error checking and validation
  • Integrated debugging capabilities

Security & Compliance

• Enhanced Tool Security: AI-driven security analysis and verification
  • Automated vulnerability detection with high confidence scoring
  • Human-in-the-loop verification for critical security decisions
  • Digital signing and integrity verification for tool distribution
• Multi-tier Sandboxing: Advanced isolation capabilities
  • Automatic risk assessment and tier selection
  • Enhanced security boundaries with cloud sandbox options
  • Improved resource management and monitoring

Fixed

• Scheduler Integration: Resolved enterprise scheduler compatibility issues
• REPL Session Management: Fixed session persistence and restoration
• Tool Review Workflow: Enhanced error handling and timeout management
• E2B Integration: Improved authentication and endpoint configuration
• Version References: Updated all version references from 0.5.0 to 0.6.0 in documentation

Breaking Changes

• Workspace Structure: New REPL crates require updated import statements
• Sandbox Architecture: Updated sandbox tier enumeration with E2B support

Dependencies

• Added: REPL system dependencies for interactive development
• Updated: Enterprise features with enhanced scheduling capabilities
• Enhanced: Tool review system with AI-powered analysis

Performance Improvements

• REPL Performance: Optimized DSL evaluation and session management
• Scheduler Throughput: Enhanced task processing with real execution support
• Tool Review Efficiency: Streamlined security analysis workflow

v0.5.0

[0.5.0] - 2025-10-14

Added

🛠️ Enhanced CLI Experience

• System Health Diagnostics: New symbi doctor command for comprehensive system health checks
  • src/commands/doctor.rs: Validates system dependencies, configuration, and runtime environment
  • Checks for required tools, permissions, and connectivity
  • Provides actionable recommendations for fixing issues
• Log Management: New symbi logs command for viewing and filtering application logs
  • src/commands/logs.rs: Real-time log streaming and filtering
  • Support for log levels, time ranges, and pattern matching
  • Integration with system logging infrastructure
• Project Scaffolding: New symbi new command for creating new agent projects
  • src/commands/new.rs: Interactive project creation with templates
  • Pre-configured project structure with best practices
  • Multiple project templates (basic, advanced, custom)
  • Automatic dependency setup and configuration
• Status Monitoring: New symbi status command for real-time system status
  • src/commands/status.rs: Display running agents, resource usage, and system health
  • Quick overview of active components and their states
• Quick Start: New symbi up command for rapid environment initialization
  • src/commands/up.rs: One-command setup for development and production
  • Automatic dependency installation and service startup
  • Health checks and validation after startup

📦 Installation & Distribution

• Automated Installation Script: New scripts/install.sh for easy setup
  • Cross-platform installation support (Linux, macOS)
  • Automatic dependency detection and installation
  • Version management and upgrade capabilities
  • Configurable installation paths and options

Improved

User Experience

• CLI Interface: Enhanced command-line interface with improved help text and error messages
  • Better command organization and discoverability
  • Consistent command structure across all operations
  • Improved error messages with actionable guidance
• README Documentation: Streamlined and updated README files across all languages
  • Simplified getting started guide
  • Clearer feature descriptions and use cases
  • Updated installation instructions
  • Better examples and quick start guides

Developer Experience

• Project Structure: Enhanced organization for better maintainability
  • Clearer separation of concerns in command modules
  • Improved code organization in src/commands/mod.rs
• Main CLI Entry Point: Updated src/main.rs with new command routing
  • Better command registration and handling
  • Enhanced error handling and logging
  • Improved startup performance

Fixed

• CLI Command Registration: Properly integrated new commands into main CLI interface
• Error Handling: Improved error messages and recovery in CLI commands
• Documentation Links: Fixed broken references in README files across all language versions

Performance Improvements

• Startup Time: Optimized CLI initialization and command loading
• Log Processing: Enhanced log streaming performance for real-time monitoring
• Status Checks: Faster system status queries and health checks

v0.4.0

[0.4.0] - 2025-08-28

Added

🧠 SLM-First Architecture (New)

• Policy-Driven Routing Engine: Intelligent routing between Small Language Models (SLMs) and Large Language Models (LLMs)
  • crates/runtime/src/routing/engine.rs: Core routing engine with SLM-first preference and LLM fallback
  • crates/runtime/src/routing/policy.rs: Configurable policy evaluation with rule-based decision logic
  • crates/runtime/src/routing/config.rs: Comprehensive routing configuration management
  • crates/runtime/src/routing/decision.rs: Route decision types and execution paths
• Task Classification System: Automatic categorization of requests for optimal model selection
  • Task-aware routing with capability matching
  • Pattern recognition and keyword analysis for task classification
• Confidence-Based Quality Control: Adaptive learning system for model performance tracking
  • crates/runtime/src/routing/confidence.rs: Confidence monitoring and threshold management
  • Real-time quality assessment with configurable confidence thresholds
  • Automatic fallback on low-confidence responses

⚡ Performance & Reliability

• Thread-Safe Operations: Full async/await support with proper concurrency handling
• Error Recovery: Graceful fallback mechanisms with exponential backoff retry logic
• Runtime Configuration: Dynamic policy updates and threshold adjustments without restart
• Comprehensive Logging: Detailed audit trail of routing decisions and performance metrics

Improved

Routing & Model Management

• Model Catalog Integration: Deep integration with existing model catalog for SLM selection
• Resource Management: Intelligent resource allocation and constraint handling
• Load Balancing: Multiple strategies for distributing requests across available models
• Scheduler Integration: Seamless integration with the existing agent scheduler

Developer Experience

• Comprehensive Testing: Complete test coverage for all routing components with mock implementations
• Documentation: Extensive design documents and implementation guides
  • docs/slm_config_design.md: SLM configuration architecture
  • docs/router_design.md: Router design and implementation guide
  • docs/unit_testing_guide.md: Testing methodology and coverage
• Configuration Validation: Enhanced validation of routing policies and model configurations

Fixed

• Module Exports: Fixed routing module structure in crates/runtime/src/routing/mod.rs
  • Added missing pub mod config; and pub mod policy; declarations
  • Added corresponding pub use statements for proper re-exports
• Task Type Updates: Replaced deprecated TaskType::TextGeneration with TaskType::CodeGeneration
  • Updated routing engine references throughout codebase
  • Fixed task type usage in test modules and policy evaluation
• Import Resolution: Resolved compilation errors in routing components
  • Updated ModelLogger constructor calls to match current API
  • Fixed import paths in test modules for proper dependency resolution
• Code Quality: Applied clippy suggestions and resolved all warnings
  • Improved code patterns and removed unused imports
  • Enhanced error handling and async operation safety

Performance Improvements

• Routing Throughput: Optimized routing decision performance with efficient policy evaluation
• Memory Efficiency: Reduced memory overhead in confidence monitoring and statistics tracking
• Async Operations: Enhanced async runtime efficiency for concurrent request handling
• Configuration Loading: Optimized configuration parsing and validation performance

Breaking Changes

• Routing API: New routing engine interface with SLM-first architecture
• Task Classification: Updated task type enumeration with CodeGeneration replacing TextGeneration
• Configuration Schema: Enhanced routing configuration structure with policy-driven settings

v0.3.1

Added

🔒 Security Enhancements

• Centralized Configuration Management: New config.rs module for secure configuration handling
  • Environment variable abstraction layer with validation
  • Multiple secret key providers (environment, file, external services)
  • Centralized configuration access patterns
• Enhanced CI/CD Security: Automated security scanning in GitHub Actions
  • Daily cargo audit vulnerability scanning
  • Clippy security lints integration
  • Secret leak detection in build pipeline

📋 API Documentation

• SwaggerUI Integration: Interactive API documentation for HTTP endpoints
  • Auto-generated OpenAPI specifications
  • Interactive API testing interface
  • Complete endpoint documentation with examples

Security Fixes

🛡️ Critical Vulnerability Resolutions

• RUSTSEC-2022-0093: Fixed ed25519-dalek Double Public Key Signing Oracle Attack
  • Updated from v1.0.1 → v2.2.0
• RUSTSEC-2024-0344: Resolved curve25519-dalek timing variability vulnerability
  • Updated from v3.2.0 → v4.1.3 (transitive dependency)
• RUSTSEC-2025-0009: Fixed ring AES panic vulnerability
  • Updated from v0.16 → v0.17.12
• Timing Attack Prevention: Implemented constant-time token comparison
  • Replaced vulnerable string comparison in authentication middleware
  • Added subtle crate for constant-time operations
  • Enhanced authentication logging and error handling

Improved

Configuration Management

• Environment Variable Security: Eliminated direct env::var usage throughout codebase
• Secret Handling: Secure configuration management with validation
• Error Handling: Enhanced configuration error reporting and validation

Authentication & Security

• Middleware Security: Updated authentication middleware to use configuration management
• Request Logging: Enhanced security logging for authentication failures
• Token Validation: Improved bearer token validation with timing attack prevention

Dependencies

Security Updates

• Updated: ed25519-dalek from v1.0.1 to v2.2.0 (critical security fix)
• Updated: reqwest from v0.11 to v0.12 (security and performance)
• Updated: ring from v0.16 to v0.17.12 (AES panic fix)
• Added: subtle v2.5 for constant-time cryptographic operations

Documentation & Tooling

• Added: utoipa and utoipa-swagger-ui for API documentation generation
• Enhanced: CI/CD security workflow with automated vulnerability scanning

v0.3.0

Added

🚀 HTTP API Server (New)

• Complete API Server: Full-featured HTTP server implementation using Axum framework
  • RESTful endpoints for agent management, execution, and monitoring
  • Authentication middleware with bearer token and JWT support
  • CORS support and comprehensive security headers
  • Request tracing and structured logging
  • Graceful shutdown with active request completion
• Agent Management API: Create, update, delete, and monitor agents via HTTP
  • Agent status tracking with real-time metrics
  • Agent execution history and performance data
  • Agent configuration updates without restart
• System Monitoring: Health checks, metrics collection, and system status endpoints
  • Real-time system resource utilization
  • Agent scheduler statistics and performance metrics
  • Comprehensive health check with component status

🧠 Advanced Context & Knowledge Management (New)

• Hierarchical Memory System: Multi-layered memory architecture for agents
  • Working Memory: Variables, active goals, attention focus for immediate processing
  • Short-term Memory: Recent experiences and temporary information
  • Long-term Memory: Persistent knowledge and learned experiences
  • Episodic Memory: Structured experience episodes with events and outcomes
  • Semantic Memory: Concept relationships and domain knowledge graphs
• Knowledge Base Operations: Comprehensive knowledge management capabilities
  • Facts: Subject-predicate-object knowledge with confidence scoring
  • Procedures: Step-by-step procedural knowledge with error handling
  • Patterns: Learned behavioral patterns with occurrence tracking
  • Knowledge Sharing: Inter-agent knowledge sharing with trust scoring
• Context Persistence: File-based and configurable storage backend
  • Automatic context archiving and retention policies
  • Compression and encryption support for sensitive data
  • Migration utilities for legacy storage formats
• Vector Database Integration: Semantic search and similarity matching
  • Qdrant integration for high-performance vector operations
  • Embedding generation and storage for context items
  • Batch operations for efficient data processing
• Context Examples: Comprehensive context_example.rs demonstration

⚡ Production-Grade Agent Scheduler (New)

• Priority-Based Scheduling: Multi-level priority queue with resource-aware scheduling
  • Configurable priority levels and scheduling algorithms
  • Resource requirements tracking and allocation
  • Load balancing with multiple strategies (round-robin, resource-based)
• Task Management: Complete lifecycle management for agent tasks
  • Task health monitoring and failure detection
  • Automatic retry logic with exponential backoff
  • Timeout handling and graceful termination
• System Monitoring: Real-time scheduler metrics and health monitoring
  • Agent performance tracking (CPU, memory, execution time)
  • System capacity monitoring and utilization alerts
  • Comprehensive scheduler statistics and dashboards
• Graceful Shutdown: Production-ready shutdown with active task completion
  • Resource cleanup and allocation tracking
  • Metrics persistence and system state preservation
  • Configurable shutdown timeouts and force termination

📊 Enhanced Documentation & Examples

• Production Examples: Real-world usage patterns and best practices
  • RAG engine integration with rag_example.rs
  • Context persistence and management workflows
  • Agent lifecycle and resource management
• API Reference: Complete HTTP API documentation with examples
  • OpenAPI-compatible endpoint specifications
  • Authentication and authorization guides
  • Integration examples for common use cases

Improved

Runtime Stability & Performance

• Memory Management: Optimized memory usage with configurable limits
• Error Handling: Enhanced error propagation and recovery mechanisms
• Async Performance: Improved async runtime efficiency and task scheduling
• Resource Utilization: Better CPU and memory resource management

Configuration & Deployment

• Feature Flags: Granular feature control for different deployment scenarios
  • http-api: HTTP server and API endpoints
  • http-input: Webhook input processing
  • vector-db: Vector database integration
  • embedding-models: Local embedding model support
• Directory Structure: Standardized data directory layout
  • Separate directories for state, logs, prompts, and vector data
  • Automatic directory creation and permission management
  • Legacy migration utilities for existing deployments

Developer Experience

• Examples: Comprehensive example implementations for all major features
• Testing: Enhanced test coverage with integration tests
• Logging: Structured logging with configurable verbosity levels
• Debugging: Improved debugging capabilities with detailed metrics

Fixed

• Scheduler Deadlocks: Resolved potential deadlock conditions in agent scheduling
• Memory Leaks: Fixed memory leaks in context management and vector operations
• Graceful Shutdown: Improved shutdown reliability under high load
• Configuration Validation: Enhanced validation of configuration parameters
• Error Recovery: Better error recovery in network and storage operations

Dependencies

• Added: Axum 0.7 for HTTP server implementation
• Added: Tower and Tower-HTTP for middleware and CORS support
• Added: Governor for rate limiting capabilities
• Added: Qdrant-client 1.14.0 for vector database operations
• Updated: Tokio async runtime optimizations
• Updated: Enhanced serialization with serde improvements

Breaking Changes

• Context API: Updated context management API with hierarchical memory model
• Scheduler Interface: New scheduler trait with enhanced lifecycle management
• Configuration Format: Updated configuration structure for directory management

Performance Improvements

• Scheduler Throughput: Up to 10x improvement in agent scheduling performance
• Memory Efficiency: 40% reduction in memory usage for large context operations
• Vector Search: Optimized vector database operations with batch processing
• HTTP Response Time: Sub-100ms response times for standard API operations

Security Enhancements

• Authentication: Multi-factor authentication support for HTTP API
• Encryption: Enhanced encryption for data at rest and in transit
• Access Control: Improved permission management for context operations
• Data Protection: Secure handling of sensitive agent data and configurations

Installation

Docker

docker pull ghcr.io/thirdkeyai/symbi:v0.3.0

Cargo (with all features)

cargo install symbi-runtime --features full

Cargo (minimal installation)

cargo install symbi-runtime --features minimal

From Source

git clone https://github.com/thirdkeyai/symbiont.git
cd symbiont
git checkout v0.3.0
cargo build --release --features full

Quick Start - HTTP API

use symbi_runtime::api::{HttpApiServer, HttpApiConfig};

let config = HttpApiConfig {
    bind_address: "0.0.0.0".to_string(),
    port: 8080,
    enable_cors: true,
    enable_tracing: true,
};

let server = HttpApiServer::new(config);
server.start().await?;

Quick Start - Context Management

use symbi_runtime::context::{StandardContextManager, ContextManagerConfig};

let config = ContextManagerConfig {
    max_contexts_in_memory: 1000,
    enable_auto_archiving: true,
    enable_vector_db: true,
    ..Default::default()
};

let context_manager = StandardContextManager::new(config, "system").await?;
let session_id = context_manager.create_session(agent_id).await?;

---

Full Changes: v0.1.2...v0.3.0