Configuration

Pulsar can be configured through environment variables, configuration files, and command-line options.

Environment Variables

API Keys

Set API keys for AI providers:

# OpenAI
export OPENAI_API_KEY="sk-..."

# Anthropic
export ANTHROPIC_API_KEY="sk-ant-..."

# Together AI (optional)
export TOGETHER_API_KEY="..."

# Other providers as needed

Ollama Configuration

Configure local Ollama instance:

# Ollama server URL (default: http://localhost:11434)
export OLLAMA_BASE_URL="http://localhost:11434"

# Model to use (can be overridden in workflows)
export OLLAMA_DEFAULT_MODEL="llama2"

Application Settings

Control Pulsar behavior:

# Log level (DEBUG, INFO, WARNING, ERROR)
export PULSAR_LOG_LEVEL="INFO"

# Default timeout for API calls (seconds)
export PULSAR_TIMEOUT="30"

# Maximum concurrent requests
export PULSAR_MAX_CONCURRENT="5"

# Cache directory for downloaded models
export PULSAR_CACHE_DIR="~/.cache/pulsar"

Configuration File

Create a .pulsar.toml or pulsar.toml file in your project directory:

[api_keys]
openai = "sk-..."
anthropic = "sk-ant-..."

[ollama]
base_url = "http://localhost:11434"
default_model = "llama2"

[app]
log_level = "INFO"
timeout = 30
max_concurrent = 5
cache_dir = "~/.cache/pulsar"

[providers.openai]
organization = "org-..."  # Optional
api_base = "https://api.openai.com/v1"  # Optional

[providers.anthropic]
api_base = "https://api.anthropic.com"  # Optional

[providers.ollama]
timeout = 60  # Override default timeout

Workflow-Specific Configuration

Override settings in individual workflows:

version: "0.1"
name: "Configured Workflow"

config:
  timeout: 60
  max_retries: 3
  log_level: "DEBUG"

agents:
  # ... agent definitions

workflow:
  # ... workflow steps

Command-Line Options

Override configuration from command line:

# Set log level
pulsar --log-level DEBUG run workflow.yml

# Set timeout
pulsar --timeout 60 run workflow.yml

# Use specific config file
pulsar --config /path/to/config.toml run workflow.yml

# Set API keys inline (not recommended for production)
pulsar --openai-key sk-... run workflow.yml

Available CLI options:

--log-level LEVEL       Set logging level
--timeout SECONDS       API call timeout
--max-concurrent N      Maximum concurrent requests
--config FILE          Configuration file path
--openai-key KEY       OpenAI API key
--anthropic-key KEY    Anthropic API key
--ollama-url URL       Ollama base URL

Provider-Specific Settings

OpenAI

[providers.openai]
api_key = "sk-..."
organization = "org-..."
api_base = "https://api.openai.com/v1"
timeout = 30
max_retries = 3

Anthropic

[providers.anthropic]
api_key = "sk-ant-..."
api_base = "https://api.anthropic.com"
timeout = 30
max_retries = 3
max_tokens = 4096

Ollama

[providers.ollama]
base_url = "http://localhost:11434"
timeout = 60
default_model = "llama2"
request_timeout = 300  # For long-running generations

Proxy Configuration

Configure HTTP proxies for API calls:

export HTTP_PROXY="http://proxy.company.com:8080"
export HTTPS_PROXY="http://proxy.company.com:8080"
export NO_PROXY="localhost,127.0.0.1,.local"

Or in configuration file:

[proxy]
http = "http://proxy.company.com:8080"
https = "http://proxy.company.com:8080"
no_proxy = "localhost,127.0.0.1,.local"

Security Considerations

API Keys

Never commit API keys to version control
Use environment variables or secure key management
Rotate keys regularly
Use read-only keys when possible

Network Security

Use HTTPS for all API communications
Configure proxies for corporate environments
Validate SSL certificates

File Permissions

Restrict access to configuration files
Use secure file permissions (600) for sensitive files

Environment Isolation

Use virtual environments for Python dependencies
Isolate different projects with separate configurations
Use Docker containers for consistent environments

Configuration Validation

Pulsar validates configuration on startup:

# Validate configuration
pulsar config validate

# Show current configuration
pulsar config show

# Check provider connectivity
pulsar config test-providers

Example Configurations

Development Environment

[app]
log_level = "DEBUG"
timeout = 60

[api_keys]
openai = "sk-test-..."

[providers.openai]
api_base = "https://api.openai.com/v1"

Production Environment

[app]
log_level = "WARNING"
timeout = 30
max_concurrent = 10

[api_keys]
openai = "sk-prod-..."
anthropic = "sk-ant-prod-..."

[providers.openai]
organization = "org-prod"
timeout = 30

[providers.anthropic]
timeout = 30

Local Development with Ollama

[app]
log_level = "INFO"

[ollama]
base_url = "http://localhost:11434"
default_model = "llama2"

[providers.ollama]
timeout = 120