Skip to content

Configuration

Configure security behavior with presets or custom settings.

Quick Start

from capiscio_sdk import SecurityConfig

# Production defaults (recommended)
config = SecurityConfig.production()

# Strict mode (fail on any issue)
config = SecurityConfig.strict()

# Custom configuration
from capiscio_sdk.config import DownstreamConfig, UpstreamConfig

config = SecurityConfig(
    downstream=DownstreamConfig(require_signatures=True),
    upstream=UpstreamConfig(validate_agent_cards=True),
    strict_mode=True,
    fail_mode="block",
)

Presets

Production

config = SecurityConfig.production()

Recommended for production deployments: - Signature verification enabled - Validation enabled - Reasonable timeouts - Standard logging

Strict

config = SecurityConfig.strict()

Maximum security: - All validations enabled - Fail on any warning - Verbose logging - No fallbacks

API Reference

capiscio_sdk.config.SecurityConfig

Bases: BaseModel

Main security configuration.

development classmethod 

development() -> SecurityConfig

Development preset - permissive.

production classmethod 

production() -> SecurityConfig

Production preset - balanced.

strict classmethod 

strict() -> SecurityConfig

Strict preset - maximum security.

from_env classmethod 

from_env() -> SecurityConfig

Load configuration from environment variables.

capiscio_sdk.config.DownstreamConfig

Bases: BaseModel

Configuration for downstream protection (agents calling you).

capiscio_sdk.config.UpstreamConfig

Bases: BaseModel

Configuration for upstream protection (calling other agents).