Skip to main content
Claudio uses environment variables to configure its behavior across all channels (Telegram, Slack, CLI). Set these in your .env file at the root of the project.

​
Quick setup

  1. Copy the example file: 
    cp .env.example .env
  2. Edit .env with your values
  3. Restart the bot to apply changes

​
Core configuration

​
WORKSPACE_PATH
string
required
The directory where Claudio executes commands and reads project context.
WORKSPACE_PATH=/path/to/claudio
This should point to your Claudio project root so it can read CLAUDE.md and access MCPs.
​
CLAUDE_CLI_PATH
string
default:"claude"
Path to the Claude Code CLI executable.
CLAUDE_CLI_PATH=claude
By default, Claudio searches for claude in your PATH. Specify a full path if needed:
CLAUDE_CLI_PATH=/usr/local/bin/claude

​
MCP and permissions

​
ALLOWED_TOOLS
string
default:"*"
Controls which tools Claude can use automatically.
ALLOWED_TOOLS=*
  • * - Allow all tools (recommended for MCPs)
  • Comma-separated list - Only allow specific tools: Read,Edit,Bash
When using MCPs, set this to * to avoid permission prompts that can’t be answered in non-interactive mode.
​
SKIP_PERMISSIONS
boolean
default:"true"
Bypass all permission checks for tools and MCPs.
SKIP_PERMISSIONS=true
This is required for Telegram and Slack bots because they run in non-interactive mode. When enabled, Claudio uses the --dangerously-skip-permissions flag.
Only enable this in trusted environments. It allows Claudio to execute any tool without confirmation.

​
Performance and limits

​
COMMAND_TIMEOUT
number
default:"1800"
Maximum execution time for commands in seconds.
COMMAND_TIMEOUT=1800
Default is 1800 seconds (30 minutes). Prevents long-running or stuck commands from blocking the bot.
​
MAX_INPUT_LENGTH
number
default:"10000"
Maximum length of user input in characters.
MAX_INPUT_LENGTH=10000
Protects against DoS attacks from extremely long messages that could consume excessive resources.
​
BUFFER_TIMEOUT
number
default:"1.5"
Time in seconds to accumulate output before sending (Telegram only).
BUFFER_TIMEOUT=1.5
Claudio buffers output to avoid flooding users with many small messages. Lower values send updates faster but create more messages.

​
Voice transcription (optional)

​
OPENAI_API_KEY
string
OpenAI API key for voice message transcription using Whisper.
OPENAI_API_KEY=sk-proj-...
Get your API key at platform.openai.com/api-keys.If not set, voice messages will not work in Telegram and Slack.
​
USE_WHISPER_API
boolean
default:"true"
Use OpenAI’s Whisper API for transcription.
USE_WHISPER_API=true
Set to false to use a local Whisper model instead (requires additional setup).
​
WHISPER_LANGUAGE
string
default:"es"
Language code for Whisper transcription.
WHISPER_LANGUAGE=es
Use en for English, es for Spanish, etc. Set to empty string for auto-detection.

​
Channel-specific variables

​
Telegram

​
TELEGRAM_BOT_TOKEN
string
required
Your Telegram bot token from @BotFather.
TELEGRAM_BOT_TOKEN=1234567890:ABCdefGHIjklMNOpqrsTUVwxyz
Get this by creating a bot with @BotFather on Telegram.

​
Slack

​
SLACK_BOT_TOKEN
string
required
Your Slack bot token (starts with xoxb-).
SLACK_BOT_TOKEN=xoxb-...
Get this from your Slack App’s “OAuth & Permissions” page.
​
SLACK_APP_TOKEN
string
required
Your Slack app-level token for Socket Mode (starts with xapp-).
SLACK_APP_TOKEN=xapp-...
Generate this in your Slack App’s “Basic Information” under “App-Level Tokens”.

​
Example configuration

.env
# Core
WORKSPACE_PATH=/Users/you/Dev/claudio
CLAUDE_CLI_PATH=claude

# Permissions (required for bots)
ALLOWED_TOOLS=*
SKIP_PERMISSIONS=true

# Security
COMMAND_TIMEOUT=1800
MAX_INPUT_LENGTH=10000
ALLOWED_USER_IDS=123456789

# Telegram
TELEGRAM_BOT_TOKEN=1234567890:ABC...

# Voice (optional)
OPENAI_API_KEY=sk-proj-...
WHISPER_LANGUAGE=es

​
Next steps

Security settings

Configure rate limiting and user authorization

MCP setup

Set up Model Context Protocol servers

Build docs developers (and LLMs) love

Get started for freeTalk to us