Skill UI

This guide details how to implement Role-Based Access Control (RBAC) for Perplexity API in enterprise environments. Since Perplexity lacks native RBAC, the solution involves building a middleware gateway. This gateway enforces granular policies, including team-specific API key usage, model restriction (e.g., limiting access to 'sonar' only), budget caps, rate limiting, and mandatory search domain filtering for compliance and security.

Orchestrate PhantomBuster cloud agents through the Composio MCP integration, automating lead generation, web scraping, usage monitoring, container tracking, CSV exports, and hCaptcha solving in a single workflow.

This skill provides comprehensive guardrails for securing Replit-hosted applications. It covers critical security policies including preventing hardcoded secrets, implementing resource usage limits (CPU/Memory), enforcing endpoint authentication, and ensuring proper deployment configuration. Use this when you need to harden a Replit app for production or audit its security posture.

This guide teaches developers how to implement robust rate limiting and handle resource quotas for applications hosted on Replit. It covers monitoring Key-Value database usage, implementing middleware for API rate limiting (using Express), and utilizing exponential backoff and request queuing patterns to ensure application stability and prevent service degradation when hitting resource limits (e.g., 429 errors).

This skill provides comprehensive patterns for data handling and integration with Retell AI. It is designed for developers building advanced voice agents, automating phone calls, or working with complex telephony systems. The guide covers SDK usage, agent listing, and robust error handling, ensuring reliable implementation of voice AI applications.

This skill provides implementation patterns for setting up and managing a multi-environment connection with Retell AI. It is essential for developers building advanced AI voice agents, integrating complex phone call automation, or utilizing robust telephony workflows. It guides users through SDK usage and handles common API error scenarios.

Serve production-grade LLM APIs with vLLM, leveraging PagedAttention, continuous batching, quantization, and tensor parallelism to maximize throughput while keeping latency, GPU usage, and scalability under control for OpenAI-compatible deployments.

A specialized methodology for reviewing APIs, configuration schemas, and interfaces to identify 'footguns'—design choices where the path of least resistance leads to insecurity. This analysis focuses on making secure usage the default, covering topics like dangerous defaults, primitive vs. semantic type safety, and insecure algorithm selection. It is crucial for architects and reviewers to prevent developer misuse at the design level.

A comprehensive guide for optimizing costs when integrating the Speak API. This skill covers usage monitoring, selecting optimal API tiers, and implementing efficient audio processing pipelines. It is essential for developers managing language learning platform operations to ensure cost efficiency while maintaining high performance.

This skill provides comprehensive observability for the Speak language learning platform. It monitors critical metrics such as API health status, assessment latency, session usage, and detailed pronunciation score distributions. It is essential for developers implementing robust, production-grade monitoring systems and managing large-scale language learning operations.

Guides using the GitHub CLI to inspect repositories, manage PRs/issues/actions, and fetch API data instead of curl/WebFetch, with tips for cloning, viewing, and automating GitHub workflows.

This skill enables AI-assisted API development workflows. It can generate type-safe API clients, comprehensive documentation, and SDKs for multiple languages (such as Python or TypeScript) directly from OpenAPI/Swagger specifications. It is essential for developers building complex integrations or needing to standardize API usage across multiple platforms.