ykachala/guardrails

Input/output safety layer for PHP LLM apps: PII redaction, prompt-injection detection, secret scanning, and content moderation as a composable pipeline.

dev-main 2026-06-01 23:22 UTC

Requires

• php: >=8.3
• psr/log: ^3.0

Requires (Dev)

• friendsofphp/php-cs-fixer: ^3.0
• phpstan/phpstan: ^2.0
• phpunit/phpunit: ^11.0

Suggests

• guzzlehttp/guzzle: Required only if you enable the optional LLM-based injection classifier
• ykachala/meter: Track the token/cost overhead added by LLM-based guards

MIT

• symfony
• security
• laravel
• moderation
• ai
• pii
• openai
• llm
• anthropic
• guardrails
• prompt-injection