npm
Toxicity model in TensorFlow.js
Safety guardrails for Reactive Agents — prompt injection detection, PII scanning, and toxicity filtering
ML-based content classifiers for AgentOS — toxicity, prompt injection, and NSFW detection via ONNX models or LLM fallback
Get sentiment and toxicity of a text.
LLM guardrails middleware — PII, injection, toxicity, schema validation, and policy engine
LangChain.js integration for open-guardrail — 215+ guards for LLM chains and agents, prompt injection, PII, toxicity & more
A React toxicity recognition wrapper capable of detecting toxic content from user's input.
Anthropic SDK adapter for open-guardrail — 215+ guards for messages input/output, prompt injection, PII, toxicity & more
This Package can detect how much toxicity persent in your text and return you toxicity Percentage in text, toxic words used in text & list of toxic word uses in the given text.
Vercel AI SDK middleware adapter for open-guardrail — 215+ guards for LLM inputs/outputs, prompt injection, PII, toxicity & more
Detect toxicity in text using React
A powerful hybrid abuse/toxicity detection module for Node.js. Combines Aho-Corasick, fuzzy matching, phonetic normalization, and TensorFlow.js AI for accurate real-time content moderation.
Profanity and abuse detection across 5 languages with severity, category, and toxicity scoring per match. Unicode-obfuscation resistant, zero dependencies, fully offline.
Express middleware adapter for open-guardrail — 215+ guards for HTTP request/response, prompt injection, PII, toxicity & more
OpenAI SDK adapter for open-guardrail — 215+ guards for chat completions input/output, prompt injection, PII, toxicity & more
Fastify plugin adapter for open-guardrail — 215+ guards for HTTP request/response, prompt injection, PII, toxicity & more
Guardrails microservice — PII detection, prompt injection defense, toxicity filtering, policy enforcement
Next.js App Router adapter for open-guardrail — 215+ guards for API routes, prompt injection, PII, toxicity & more
Hono middleware adapter for open-guardrail — 215+ guards for edge runtimes, prompt injection, PII, toxicity & more
A comprehensive JavaScript library for content moderation, including profanity filtering, sentiment analysis, and toxicity detection. Leveraging advanced algorithms and external APIs, TextModerate provides developers with tools to create safer and more po
350 built-in guards for LLM safety: 26 PII regions, prompt injection, toxicity, bias, agent safety, hallucination, content safety, compliance, format validation, Korean ISMS-P/PIPA, GDPR, EU AI Act
A TypeScript implementation of decompression calculation algorithms for scuba diving, featuring Bühlmann ZH-L16C algorithm with gradient factors, gas management, and oxygen toxicity tracking.
Model Context Protocol server for the Sonny Labs AI firewall. Run it inside your agentic client (Claude Desktop, Cursor, Claude Code) to scan prompts and outputs for injection / PII / toxicity, manage API keys, and scaffold the firewall into your codebase
Traits for ORM layer
ML and pattern-based classifiers for toxicity, PII, and prompt injection detection
A platform for building web applications with Rust
A dive decompression models library (Buhlmann ZH-L 16C)
YAML-based policy engine for LLM safety rules, triggers, and actions
Hash-chained audit trail and Prometheus metrics for CheckStream
Shared lint analysis core for modum
ONNX-runtime-backed scanners for llm-guard. Catches paraphrased / novel prompt-injection attacks the rules tier can't. CPU by default; CUDA / CoreML / DirectML opt-in.
Reusable AI assistant library for local LLM integration (Ollama, LM Studio, etc.)
no toxic names anymore
Lightweight client for Toxiproxy
VIL Content Guardrails Engine — PII detection, toxicity scoring, custom rules (H07)
As prompted by the user, it can list all plants starting with a given letter of the alphabet as well as provide further details about any plant listed.
Graphical User Interface for Lazar Toxicity Predictions
QMRF and QPRF reporting for OpenTox ruby module and Lazar Toxicity Predictions
Provides LLM-as-judge and code-based evaluators for scoring LLM outputs, with built-in templates for hallucination, relevance, and toxicity detection.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.