ai-collaboration-standards
Prevents AI hallucination and ensures evidence-based, verifiable outputs when analyzing code, reviewing technical documents, or providing recommendations.
Discover reusable agent skills, browse implementation details, and find the right skill for your workflow.
557 skills found
Prevents AI hallucination and ensures evidence-based, verifiable outputs when analyzing code, reviewing technical documents, or providing recommendations.
Optimize agent performance and token usage through advanced context compression, structured summarization, and task-oriented state management for long-running sessions.
Master workflow controller for Lovable-style, AI-driven development. Instantly generates premium, multi-page, animated applications by routing to specialized sub-agents. No prompts needed—just build.
Multi-phase feature development workflow for complex tasks using research, planning, implementation, and review gates.
Validates Excel exports for Customer Feedback Analyzer with 7 specific view sheets, 36 columns, and precise color-coded formatting. Ensures zero errors in customer-facing deliverables.
Rigorous research skill that enforces source verification via WebFetch and content analysis to prevent hallucinated citations.
Fast lookup for SaaS finance metrics, formulas, and benchmarks. Optimize your financial analysis with quick access to definitions, decision frameworks, and red flag indicators.
Epistemic safety analysis for JSON data in prompts to prevent LLM hallucinations and reasoning errors when handling incomplete or large-scale datasets.
Autonomous research specialist for verified information gathering, source evaluation, and structured synthesis.
An end-to-end video processing pipeline that transforms raw recordings into transcripts, key insights, short clips, and polished articles.
A testing skill designed to verify the functionality of the Skillet CLI by performing basic tasks and confirming completion.
AI-powered browser automation server for web interaction, data extraction, and research using the Model Context Protocol.