⚠️ This validator has moved

This validator now lives in the Guardrails Hub monorepo. This repository is archived and no longer maintained — please open issues and pull requests on the monorepo instead.

pip install guardrails-ai-ban-list

from guardrails import Guard
from guardrails_ai.ban_list import BanList

guard = Guard().use(BanList)

The registered validator name is unchanged, so existing guards keep working.

---

Overview

Developed by	Guardrails AI
Date of development	Aug 16, 2024
Validator type	Data Leakage
Blog
License	Apache 2
Input/Output	Input

Description

Validates that output does not have banned words, using fuzzy search. Useful for preventing internal codenames from leaking.

Intended Use

Requirements

• Dependencies:

  • guardrails-ai>=0.4.0
  • fuzzysearch

• Foundation model access keys:

  • OPENAI_API_KEY

Installation

$ guardrails hub install hub://guardrails/ban_list

Usage Examples

Validating string output via Python

In this example, we apply the validator to a string output generated by an LLM.

# Import Guard and Validator
from guardrails.hub import BanList
from guardrails import Guard

# Setup Guard
guard = Guard().use(
    BanList(banned_words=['codename','athena'])
)

guard.validate("Hello world! I really like Python.")  # Validator passes
guard.validate("I am working on a project with the code name A T H E N A")  # Validator fails

__init__(self, on_fail="noop")

Initializes a new instance of the BanList class.

Parameters

• banned_words (List[str]): A list of banned words to check for in output.
• max_l_dist (int): Maximum Levenshtein distance for fuzzy search. Defaults to 1.
• on_fail (str, Callable): The policy to enact when a validator fails. If str, must be one of reask, fix, filter, refrain, noop, exception or fix_reask. Otherwise, must be a function that is called when the validator fails.

validate(self, value, metadata) -> ValidationResult

Validates the given value using the rules defined in this validator, relying on the metadata provided to customize the validation process. This method is automatically invoked by guard.parse(...), ensuring the validation logic is applied to the input data.

Note:

1. This method should not be called directly by the user. Instead, invoke guard.parse(...) where this method will be called internally for each associated Validator.
2. When invoking guard.parse(...), ensure to pass the appropriate metadata dictionary that includes keys and values required by this validator. If guard is associated with multiple validators, combine all necessary metadata into a single dictionary.

Parameters

• value (Any): The input value to validate.
• metadata (dict): A dictionary containing metadata required for validation. Keys and values must match the expectations of this validator.

  Key	Type	Description	Default
  key1	String	Description of key1's role.	N/A