Guardrails

A minimalist TypeScript middleware framework for building AI agents with composable guards, model routing, memory, testing, and observability.

A set of small, single-purpose libraries that improve reliability in production LLM agents through tool validation, network controls, trace tests, context fitting, structured output checks, and usage budgets.

An MCP server for AI agents providing cryptographic activity provenance, chain integrity checks, agent identity verification, and reputation tools.

An agent identity and trust platform providing signed credentials, registry discovery, reputation scoring, and verification tools.

AIActGuard provides EU AI Act compliance middleware for agentic AI, with audit trails, risk classification, and human-approval gates integrated into existing agent frameworks, plus tools for drafting supporting documentation.

An AI-native file format and toolchain for carrying provenance, evidence, and verification metadata with files.

APort Agent Guardrails evaluates tool actions before execution and applies configurable authorization policies for agent frameworks, including OpenClaw.

A local credential broker for AI agents that stores credentials in an encrypted vault and injects them through a local HTTPS proxy, keeping raw secrets away from agent processes.

An OpenClaw-based multi-agent workforce framework with shared team memory, goal decomposition, fleet learning, task management, and approval workflows.

A pre-transaction safety hook for AI agents that assesses crypto contract and token risks using security-oracle data and bytecode analysis.

A self-hosted runtime authority for AI agents that enforces budgets, risk limits, and tool-action policies before execution through a concurrency-safe reservation lifecycle.

Guardrails is a Python framework for building reliable AI applications, offering Input/Output Guards to detect and mitigate risks, along with structured data generation from large language models (LLMs)

The text describes 'guidance,' a programming paradigm that enhances control and efficiency in model generation by allowing for constraints like regex and CFGs, integrating stateful control, and offering a simplified interface for complex generation scenarios

The open-source Hermes Plant MCP server connects agents to action safety and commerce assurance services for shell, Git, SQL, deployment, and x402 workflows.

An open-source trust registry that aggregates public evidence about AI agents and MCP servers, with project comparisons, score history, and machine-readable data.

jes is an open-source guardrails library for checking agent inputs, retrieved content, tool calls, tool results, and responses for prompt injection, jailbreaks, and sensitive-data leaks.

A Kubernetes operator for deploying and managing AI agent runtimes, with bounded self-healing that keeps LLMs out of the cluster's workload write path.

NeMo Guardrails is an open-source toolkit facilitating the integration of programmable guardrails, essential for steering and safeguarding AI agents' conversational outputs, into large language model-based applications