CC-Telegram-Bridge: Run Claude Code and Codex CLI in Telegram

CC-Telegram-Bridge is an innovative bridging tool that enables users to directly run Claude Code and Codex CLI within the Telegram chat interface. It supports session recovery, multi-bot instance isolation, Agent Bus workflow delegation, voice input, and real-time streaming responses, bringing professional AI programming assistants into instant messaging scenarios.

Its core value lies in democratizing professional AI programming capabilities to broader scenarios, allowing activities like code review, rapid prototype validation, and team collaborative programming to be completed anytime, anywhere via a familiar chat interface.

Native CLI Encapsulation Layer

CC-Telegram-Bridge deeply integrates the native CLI capabilities of Claude Code and Codex CLI through process management and standard I/O redirection, reproducing the full CLI interaction experience in the Telegram Bot architecture. This includes complete command parsing/execution, file system operations/code editing, tool call chains, and multi-round dialogue context maintenance.

Session Persistence & Recovery

To address the stateless nature of Telegram Bot architecture, it implements an intelligent session management system:

• Session state serialization: Stores each user's dialogue context, file modification history, and tool execution results for cross-device recovery.
• Breakpoint resume: Sessions can be accurately restored from the last state even if the connection is interrupted or devices are switched.
• Versioned session management: Supports multiple parallel sessions for flexible switching between different project/task contexts.

Multi-Bot Instance Isolation

For team scenarios, it provides enterprise-level isolation:

• Process-level isolation: Each user/team's Bot instance runs in an independent process.
• Data isolation: Strict separation of code bases, configuration files, and environment variables.
• Permission control: Role-based access control (RBAC) for configuring user operation permissions.

The project's most innovative feature is the Agent Bus architecture, enabling complex multi-agent collaboration:

Delegation Mode

When users propose complex requirements, the main agent can intelligently decompose tasks and delegate to specialized sub-agents. For example, a request to 'build a complete user authentication system' may be split into database design, backend API, frontend UI, and testing sub-agents.

Fan-out Mode

For parallelizable tasks, Agent Bus supports launching multiple sub-agents simultaneously to handle different subtasks and aggregate results, significantly improving efficiency.

Crew Workflows

Drawing on human team collaboration, Crew Workflows allow multiple agents to take on roles like architect, developer, and reviewer, completing large projects via structured collaboration with clear responsibility boundaries and interaction protocols.

Voice Input Support

Integrates speech-to-text functionality for mobile convenience, allowing users to quickly describe requirements via voice, which is converted into structured instructions.

Streaming Response & Real-Time Feedback

Supports true streaming output, with AI thinking processes, code generation progress, and tool execution status displayed in real time, enhancing user experience and trust.

Rich Media Output

Supports multiple output formats: syntax-highlighted code blocks, visual code difference comparisons, command line execution results, and Mermaid flowcharts/architecture diagrams.

Code Safety

• All code runs in a sandbox environment to prevent malicious code from affecting the host system.
• Sensitive operations (e.g., deleting files, executing system commands) require explicit user confirmation.
• Supports code review mode, showing all changes to users before execution.

Data Privacy

• User code and dialogue data can be stored with end-to-end encryption.
• Supports private deployment to keep sensitive data within the local environment.
• Provides detailed audit logs for all key operations.

Mobile Development & Rapid Prototyping

Developers can quickly validate ideas on the go (e.g., commuting, meeting gaps) by forwarding code snippets to the Bot for explanation, rewriting, or test case generation.

Team Collaboration & Code Review

In team groups, the Bot can participate in technical discussions: analyze shared code fragments, propose improvements, generate code review checklists, and automatically create fix PRs.

Education & Knowledge Sharing

Tutors can use the Bot to demonstrate code concepts in group teaching, while students can ask real-time questions and get personalized code examples.

Operations & Monitoring

Combined with tool call capabilities, it can query server status, analyze log files, execute standardized troubleshooting processes, and generate operation reports.

Enhanced Collaboration Features

• Real-time multi-person collaborative editing.
• Voice/video call integration.
• Screen sharing and remote debugging.

Intelligent Project Management

• Automatic task decomposition and progress tracking.
• Integration with Jira, GitHub Issues, etc.
• Intelligent workload estimation.

Enterprise-Level Features

• SSO single sign-on.
• Fine-grained permission management.
• Audit and compliance reports.
• SLA guarantee and monitoring.

CC-Telegram-Bridge represents an important evolution direction of AI-assisted programming tools—breaking platform boundaries to make powerful AI capabilities ubiquitous. It is not just a technical tool but an exploration of a new work style: integrating AI programming assistants into daily communication scenarios to make technical collaboration more natural and efficient.

For developers pursuing flexible work styles, technical teams needing mobile office support, and organizations hoping to lower the threshold for using AI tools, CC-Telegram-Bridge provides an innovative and practical solution. As AI programming capabilities continue to evolve, this 'AI-as-a-service' integration mode via instant messaging is expected to become a new industry standard.