chenhg5/cc-connect
Cc Connect
cc-connect is an AI agent messaging bridge and session manager that connects local AI coding agents to third-party messaging platforms. It acts as a multimodal AI chat relay and a OneBot protocol gateway, allowing users to control local AI agents remotely via a variety of chat interfaces.
The project distinguishes itself by providing a remote AI agent controller that enables the management of agents through slash commands and a web management dashboard. It supports multi-tenant project orchestration and session-based context isolation, ensuring that independent conversation threads are maintained per user or group to prevent context drift.
The system covers a broad range of capability areas, including multimodal content forwarding for voice and image processing, the execution of local shell commands, and natural language task scheduling. It also provides security features such as OS-level process isolation, role-based access control, and manual tool approval workflows. Integration support extends to multiple platforms including Discord, Slack, WeChat, Telegram, and others.
The application can be installed as a system-level background daemon using systemd or launchd for persistent operation.
Features
- AI Messaging Bridges - Acts as a specialized bridge connecting local AI agents to various third-party messaging platforms.
- Group Context Isolation - Creates independent sessions for each root message in group chats to prevent cross-conversation interference.
- Agent Interfaces - Defines standardized session and agent interfaces to support diverse AI assistants at runtime.
- Agent Memory Persistence - Maintains long-term agent context by reading and writing instruction files via chat commands.
- Agent Session Management - Manages isolated conversation contexts, model providers, and working directories for multiple AI agents.
- Remote Control Layers - Provides a remote control layer to manage local AI agents via slash commands and a web dashboard.
- AI Coding Assistant Integrations - Bridges AI coding assistants like Claude Code or Cursor to third-party chat applications.
- AI Provider Integrations - Maps API endpoints and model identifiers to establish connectivity with various third-party AI providers.
- Conversation Threads - Maintains dedicated conversation threads for individual sessions to organize follow-up interactions.
- Session Management - Maintains user-specific context through independent sessions and inactivity-based rotation.
- Multi-Project Process Managers - Manages multiple independent projects in a single process, each with its own unique agent and platform mappings.
- Agent-Platform Mapping - Orchestrates multiple independent projects, mapping specific agents to their respective messaging platforms.
- Chat Session Managers - Manages independent conversation sessions for every participant within a single group chat.
- Chat Protocol Bridges - Bridges third-party messaging accounts to local AI agents, allowing interaction without installing enterprise software.
- WebSocket Proxy Tunnels - Implements WebSocket tunnels to route network traffic from external messaging platforms to local agents across restricted boundaries.
- Message Routing - Uses the OneBot protocol to standardize message routing between different platforms and AI agents.
- Messaging Platform Integrations - Provides adapters to connect local AI agents to various external social messaging services.
- Remote Agent Task Execution - Enables remote execution of AI agent tasks by connecting local assistants to messaging platforms.
- Chatbot Gateways - Routes interactions between local AI agents and messaging services like Slack, Discord, and Telegram.
- Conversational Context Isolation - Ensures independent conversation threads per user or group to maintain context purity.
- Agent Extensibility Frameworks - Provides a standard interface that allows integrating new AI assistants into the bridge at runtime.
- Skill Installation - Enables discovering and adding new capability packages from an open ecosystem to extend agent functionality.
- In-Chat Instruction Updates - Allows updating agent memory and instruction files directly from the chat interface to persist behavioral changes.
- In-Chat Model Controls - The AI bridge allows users to switch models and reasoning intensity using slash commands directly within the chat interface.
- Tool Execution Approvals - Provides a toggle between automatic tool execution and manual confirmation for every agent action.
- Dynamic Model Switching - Allows users to update AI intelligence or switch providers on the fly via chat commands.
- Context Compression - Reduces token consumption by compacting the active conversation context window.
- Session Rotation Strategies - Rotates sessions after periods of inactivity to prevent context drift in long-running AI interactions.
- Voice Synthesis - Synthesizes text responses into speech and delivers them as audio messages to the user.
- Multi-Agent Orchestration - Binds multiple AI agents to a single group chat for collaborative problem solving.
- Multimodal AI Conversions - Transforms data between text, speech, and image modalities to facilitate communication between chat platforms and AI agents.
- Multimodal Voice Integrations - Integrates speech-to-text and text-to-speech systems to enable multimodal interaction via voice and screenshots.
- Real-Time Text Streaming - Implements real-time token streaming to update messages in-place as the agent produces output.
- User-Defined Skill Discovery - Discovers and invokes user-defined skill files from agent directories to extend assistant capabilities.
- Natural Language Cron Schedulers - Translates natural language descriptions into recurring jobs for automated periodic reports and summaries.
- Development Workflow Automation - Automates developer workflows by scheduling natural language tasks and executing local shell commands.
- Runtime Setting Controls - Modifies system behavior and display modes in real-time via a chat command interface or web dashboard.
- Shell Command Execution - Executes system shell commands directly via custom configuration or chat-based shortcuts.
- Slash Command Interfaces - Enables the creation of parameterized prompt templates via slash commands to standardize frequent requests.
- Session Interaction APIs - Provides a REST and WebSocket server for external adapters to interact with AI sessions.
- Protocol Gateways - Implements a gateway that routes interactions between AI assistants and chat applications using the OneBot standard.
- Remote Interface Relays - Relays voice, images, and files between external messaging platforms and local AI agent interfaces.
- Voice-to-Text Messaging Integrations - Converts incoming voice messages from chat platforms into text for processing by AI agents.
- Process User Isolation - Spawns agent processes under specific Unix user identities to enforce filesystem and permission boundaries.
- Command Restrictions - Restricts high-risk shell commands and directory changes to a list of authorized users.
- Role-Based Access Control - Manages bridge interaction through authorized user ID lists and per-user rate limits.
- Event-Driven Hook Systems - Triggers shell commands or webhooks based on system events like message delivery.
- Session Lifecycle Hooks - Triggers custom shell commands or webhooks automatically upon session creation, message delivery, or session termination.
- Web Management Dashboards - Provides a browser-based interface for managing projects, platforms, and providers as an alternative to config files.
- Messaging Platform Adapters - Allows adding support for additional chat applications by implementing and registering a platform interface.
- Herramientas de desarrollo - Bridge for connecting coding agents to messaging platforms.
