Skill UI

Adds x402 payment execution to AI agents with per-task budgets, spending controls, and non-custodial wallets. Supports Base via agentwallet-sdk and X Layer via OKX Payments for safe API purchases and agent-to-agent payments.

Trackio is a comprehensive library for logging, visualizing, and managing ML experiment metrics and alerts. It allows developers to use a Python API to log metrics and set diagnostic alerts (INFO/WARN/ERROR) during training. Furthermore, a dedicated CLI enables metric retrieval and dashboard viewing, syncing data to Hugging Face Spaces for persistent, remote monitoring, making it ideal for autonomous ML agent workflows.

This framework implements robust input and output validation guardrails for LLM-powered applications. It prevents prompt injection, detects and redacts PII, filters toxic content, and validates model outputs against defined schemas. Use it when deploying AI agents that handle sensitive data or require strict adherence to corporate policies, ensuring a secure and trustworthy AI interaction layer.

The Safety Guard skill is designed to prevent destructive actions when working in production systems or when autonomous agents are executing tasks. It offers three protection modes (Careful, Freeze, and Guard) to monitor dangerous commands, lock specific directories, and ensure code changes are confined to intended areas, enhancing system stability and security.

The Santa Method is a multi-agent adversarial verification framework designed to improve output quality by requiring two independent reviewers (Agent B and Agent C) to pass the content before publication. It eliminates single-agent bias and systematically improves outputs through iterative refinement loops until convergence or manual escalation.

This tool assesses whether coding agents adhere to defined skills, rules, or agent definitions. It automatically generates scenarios with varying prompt strictness, executes the agent, and generates a comprehensive report detailing the compliance rate of tool calls against predefined workflow specifications using advanced LLM classification.

Assesses AI/ML systems for prompt injection, jailbreak, model inversion, data poisoning, and agent tool abuse using MITRE ATLAS mapping, signature detection, and scoring metrics before deployment.

This harness transforms standard code environments into fully persistent, autonomous agent systems. It leverages native features like scheduled tasks (Crons), remote dispatching, and structured memory storage (MCP) to enable continuous, self-directing automation. Ideal for building AI assistants that require long-term context, recurring workflows, and complex, multi-step operations, eliminating the need for external agent frameworks like AutoGPT.

A comprehensive multi-agent feedback loop inspired by Generative Adversarial Networks (GANs), designed to build complex, production-ready applications. It separates generation from evaluation, using a 'ruthlessly strict' evaluator (QA) to drive continuous iteration. This process ensures high design quality, robust functionality, and originality, overcoming the limitations of single-agent self-critique. Ideal for full-stack feature development and complex product specification implementation.

AgentFlow transforms your existing Kanban board (Asana, GitHub, Linear) into a self-managing, autonomous AI development pipeline. It orchestrates multiple AI agents through defined stages (Backlog to Done), enforcing deterministic quality gates, performing adversarial reviews, and tracking costs—all visible within your existing project management UI. It provides crash-proof execution and full observability for complex software development lifecycles.

AgentPhone is an API-first platform designed for developers building AI agents with advanced telephony capabilities. It allows agents to manage phone numbers, make/receive voice calls, and handle SMS messages. Developers can utilize hosted mode for simple AI interactions or webhook mode for custom, complex logic, making it ideal for building sophisticated customer support bots or virtual assistants.

Guides calling Feishu IM APIs to retrieve chat histories, thread replies, cross-chat searches, and media/file downloads while handling pagination, time filters, and permission constraints so agents can answer questions about past conversations.