Skip to main content

bellwether init

Create a default configuration file.

Synopsis​

bellwether init [options]

Description​

The init command creates an bellwether.yaml configuration file in the current directory with sensible defaults. This file can be customized to configure LLM providers, interview settings, and output options.

Options​

OptionDescriptionDefault
-f, --forceOverwrite existing config filefalse

Examples​

Create Default Config​

bellwether init

Creates bellwether.yaml:

version: 1

# LLM Provider Configuration
llm:
  provider: openai          # openai, anthropic, or ollama
  model: gpt-4o             # Optional: override default model

# Interview Settings
interview:
  maxQuestionsPerTool: 3    # Questions per tool (1-20)
  timeout: 30000            # Tool call timeout in ms
  personas:                 # Personas to use (optional)
    - technical_writer
    - security_tester

# Output Settings
output:
  format: markdown          # markdown, json, or both
  directory: .              # Output directory

Overwrite Existing Config​

bellwether init --force

Configuration Options​

LLM Section​

llm:
  provider: openai      # Required: openai, anthropic, or ollama
  model: gpt-4o-mini    # Optional: specific model
  baseUrl: http://...   # Optional: custom API endpoint

Interview Section​

interview:
  maxQuestionsPerTool: 5    # 1-20 questions per tool
  timeout: 60000            # Timeout per tool call (ms)
  personas:                 # Which personas to use
    - technical_writer
    - security_tester
    - qa_engineer
    - novice_user
  personaFiles:             # Custom persona files
    - ./personas/custom.persona.yaml

Output Section​

output:
  format: markdown      # markdown, json, sarif, junit
  directory: ./docs     # Where to write output files
  agentsFileName: AGENTS.md

See Also​