ductor 0.7.0

Control Claude Code and Codex CLI from Telegram. Live streaming, sessions, cron jobs, webhooks, Docker sandboxing.

Claude Code, Codex CLI, and Gemini CLI as your Telegram assistant.
Named sessions. Persistent memory. Scheduled tasks. Live streaming. Docker sandboxing.
Uses only official CLIs. Nothing spoofed, nothing proxied.

Quick start · Features · How it works · Commands · Docs · Contributing

---

Use your Claude Max, GPT Pro, or Gemini Pro subscription with ductor. Control your coding agents via Telegram -- automations, cron jobs, named sessions, and more.

ductor runs on your machine, uses your existing CLI authentication, and keeps state in plain JSON/Markdown under ~/.ductor/.

Quick start

pipx install ductor
ductor

The onboarding wizard handles CLI checks, Telegram setup, timezone, optional Docker, and optional background service install.

Requirements: Python 3.11+, at least one CLI installed (claude, codex, or gemini), a Telegram Bot Token from @BotFather.

Detailed setup: docs/installation.md

Why ductor?

ductor executes the real provider CLIs as subprocesses. No API proxying, no spoofing.

Other projects manipulate SDKs or patch CLIs and risk violating provider terms of service. ductor simply runs the official CLI binaries as if you typed the command in your terminal. Nothing more.

• Official CLIs only (claude, codex, gemini)
• Rule files are plain Markdown (CLAUDE.md, AGENTS.md, GEMINI.md)
• Memory is one Markdown file (memory_system/MAINMEMORY.md)
• All state is JSON (sessions.json, cron_jobs.json, webhooks.json)

Features

Core

• Real-time streaming with live Telegram edits
• Provider/model switching with /model (sessions are preserved per provider)
• @model directives for inline provider targeting
• Inline callback buttons, queue tracking with per-message cancel
• Persistent memory in plain Markdown

Named sessions

Run tasks in the background while you keep chatting. Each session gets a unique name and supports follow-ups:

/session Fix the login bug              -> starts "firmowl" on default provider
/session @codex Refactor the parser     -> starts "pureray" on Codex
/session @opus Analyze the architecture -> starts "goldfly" on Claude (opus)
/session @flash Check the logs          -> starts "slimelk" on Gemini (flash)

@firmowl Also check the tests           -> foreground follow-up
/session @firmowl Add error handling     -> background follow-up

/sessions                                -> list/manage active sessions

@model shortcuts resolve the provider automatically (@opus = Claude, @flash = Gemini, @codex = Codex).

Automation

• Cron jobs: in-process scheduler with timezone support, per-job overrides, quiet hours
• Webhooks: wake (inject into active chat) and cron_task (isolated task run) modes
• Heartbeat: proactive checks in active sessions with cooldown + quiet hours
• Config hot-reload: safe fields update without restart (mtime-based watcher)

Infrastructure

• Service manager: Linux (systemd), macOS (launchd), Windows (Task Scheduler)
• Docker sandbox: sidecar container with configurable host mounts
• Auto-onboarding: interactive setup wizard on first run
• Cross-tool skill sync: shared skills across ~/.claude/, ~/.codex/, ~/.gemini/

How it works

graph LR
    A[You on Telegram] --> B[aiogram Middleware]
    B --> C[Orchestrator]
    C --> D[CLIService]
    D --> E[claude]
    D --> F[codex]
    D --> G[gemini]
    E & F & G --> H[Streamed response]
    H --> A

    C --> I[Background Systems]
    I --> J[Cron / Webhooks / Heartbeat]
    I --> K[Named Sessions]

The orchestrator routes messages through command dispatch, directive parsing, and conversation flows. Background systems (cron, webhooks, heartbeat, named sessions, config reload, model caches) run as in-process asyncio tasks.

Session behavior:

• Sessions are chat-scoped and provider-isolated
• /new resets only the active provider bucket
• Switching providers preserves each provider's session context

Telegram commands

Command	Description
/session <prompt>	Run named background session
/sessions	View/manage active sessions
/model	Interactive model/provider selector
/new	Reset active provider session
/stop	Abort active run
/status	Session/provider/auth status
/memory	Show persistent memory
/cron	Interactive cron management
/showfiles	Browse ~/.ductor/
/diagnose	Runtime diagnostics
/upgrade	Check/apply updates
/info	Version + links

CLI commands

ductor                  # Start bot (auto-onboarding if needed)
ductor stop             # Stop bot
ductor restart          # Restart bot
ductor upgrade          # Upgrade and restart
ductor status           # Runtime status

ductor service install  # Install as background service
ductor service logs     # View service logs

ductor docker enable    # Enable Docker sandbox
ductor docker rebuild   # Rebuild sandbox container
ductor docker mount /path  # Add host mount

ductor api enable       # Enable WebSocket API (beta)

Full CLI reference: docs/modules/setup_wizard.md

Workspace layout

~/.ductor/
  config/config.json        # Bot configuration
  sessions.json             # Chat session state
  named_sessions.json       # Named background sessions
  cron_jobs.json            # Scheduled tasks
  webhooks.json             # Webhook definitions
  CLAUDE.md / AGENTS.md / GEMINI.md  # Rule files
  logs/agent.log
  workspace/
    memory_system/MAINMEMORY.md      # Persistent memory
    cron_tasks/ skills/ tools/       # Task scripts, skills, tool scripts
    telegram_files/ output_to_user/  # File I/O directories

Full config reference: docs/config.md

Documentation

Doc	Content
Developer Quickstart	Fastest path for contributors
Architecture	Startup, routing, streaming, callbacks
Configuration	Config schema and merge behavior
Automation	Cron, webhooks, heartbeat setup
Module docs	Per-module deep dives (24 modules)

Disclaimer

ductor runs official provider CLIs and does not impersonate provider clients. Validate your own compliance requirements before unattended automation.

• Anthropic Terms
• OpenAI Terms
• Google Terms

Contributing

git clone https://github.com/PleasePrompto/ductor.git
cd ductor
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pytest && ruff format . && ruff check . && mypy ductor_bot

Zero warnings, zero errors.

License

MIT