Configuration

Kader can be configured through environment variables, YAML files, and the .kader directory in your home folder.

Environment Variables

Variable	Description	Required
KADER_DIR	Kader config directory (default: ~/.kader)	No
OLLAMA_API_KEY	Ollama Cloud API key (get from https://ollama.com/settings)	For Ollama Cloud
GEMINI_API_KEY	Google Gemini API key	For Google Provider
MISTRAL_API_KEY	Mistral API key	For Mistral Provider
ANTHROPIC_API_KEY	Anthropic API key	For Anthropic Provider
OPENAI_API_KEY	OpenAI API key	For OpenAI Provider
MOONSHOT_API_KEY	Moonshot AI key	For Moonshot Provider
ZAI_API_KEY	Z.ai (GLM) API key	For Z.ai Provider
OPENROUTER_API_KEY	OpenRouter key	For OpenRouter Provider
OPENCODE_API_KEY	OpenCode API key	For OpenCode Provider
GROQ_API_KEY	Groq API key	For Groq Provider

API keys can be set interactively using the /connect command in the CLI, or manually by editing ~/.kader/.env.

.kader Directory

When the kader module is imported for the first time, it automatically creates a .kader directory in your home directory:

~/.kader/
├── .env                   # Environment variables
├── settings.json          # User preferences (model/provider)
├── memory/               # Memory and session storage
│   └── sessions/         # Saved conversation sessions
├── skills/               # User-level skills
├── commands/             # User-level special commands
├── custom/
│   ├── callbacks/        # User-level callbacks
│   └── tools/            # User-level custom tools
└── KADER.md              # Agent instructions file

YAML Agent Configuration

BaseAgent supports loading configuration from YAML files:

# agent.yaml
name: ConfiguredAgent
system_prompt: "You are a helpful coding assistant."
tools:
  - read_file
  - write_file
  - todo_tool
provider:
  model: llama3.2
  provider: ollama
persistence: true
retry_attempts: 3
interrupt_before_tool: true

agent = BaseAgent.from_yaml("agent.yaml")

Provider Configuration Format

Ollama (Local)

provider:
  provider: ollama
  model: llama3.2
  base_url: "http://localhost:11434"
  timeout: 120

Ollama (Cloud)

provider:
  provider: ollama
  model: minimax-m2.5
  base_url: "https://ollama.com"
  # api_key: "your-ollama-api-key"  # Or set OLLAMA_API_KEY env var

Google Gemini

provider:
  provider: google
  model: gemini-2.0-flash
  temperature: 0.7
  max_tokens: 2048

Anthropic

provider:
  provider: anthropic
  model: claude-3-5-sonnet-20241022
  temperature: 0.7

OpenAI-Compatible

provider:
  provider: openai
  model: gpt-4o
  base_url: "https://api.openai.com/v1"
  api_key: "your-openai-key"

For Groq, OpenRouter, etc., use the appropriate base_url.

Session Storage

Sessions are saved to ~/.kader/memory/sessions/ and include: - Conversation history - Agent state - Tool execution logs - Sub-agent contexts (for aggregated context)

Settings

User preferences are stored in ~/.kader/settings.json, created automatically on first run:

{
  "main-agent-provider": "ollama",
  "sub-agent-provider": "ollama",
  "main-agent-model": "glm-5:cloud",
  "sub-agent-model": "glm-5:cloud",
  "auto-update": false,
  "callbacks": [],
  "tools": [],
  "subagents": []
}

Field	Description	Default
main-agent-provider	LLM provider for the planner agent	ollama
sub-agent-provider	LLM provider for executor sub-agents	ollama
main-agent-model	Model name for the planner agent	glm-5:cloud
sub-agent-model	Model name for executor sub-agents	glm-5:cloud
auto-update	Automatically update Kader on startup	false
callbacks	List of user-level callbacks to enable	[]
tools	List of user-level custom tools to enable	[]
subagents	List of user-level subagents to enable	[]

Settings are updated automatically when switching models via the /models CLI command. You can also edit the file directly.

Callbacks Configuration

{
  "callbacks": [
    {"name": "my_callback", "enabled": "true"},
    {"name": "other_callback", "enabled": "false"}
  ]
}

• name: The filename (without .py extension) in ~/.kader/custom/callbacks/
• enabled: "true" to enable, "false" to disable

Tools Configuration

{
  "tools": [
    {"name": "my_tool", "enabled": "true", "agent": "executor"},
    {"name": "other_tool", "enabled": "false", "agent": "both"}
  ]
}

• name: The filename (without .py extension) in ~/.kader/custom/tools/, or module.ClassName
• enabled: "true" to enable, "false" to disable
• agent: "planner", "executor", or "both" (which agents can use the tool)

Subagents Configuration

User-level subagents in ~/.kader/subagents/ are gated by the subagents field in settings:

{
  "subagents": [
    {"name": "code-reviewer", "enabled": "true"},
    {"name": "research-agent", "enabled": "false"}
  ]
}

• name: Matches the subagent YAML name field (or the YAML file stem)
• enabled: "true" to enable, "false" to disable
• Subagents not listed in settings are skipped
• Project-level subagents in ./.kader/subagents/ are always enabled regardless of settings

Auto-Update

When auto-update is set to true, Kader will automatically check for and install updates on every startup. The update is performed silently using uv tool upgrade kader.

You can also manually check for updates using the /update command. If a newer version is available, it will upgrade Kader and restart the CLI. If you're already on the latest version, it will display a confirmation message.