ws-protocol
ws-protocol is a technical specification outlining the protocol version, message envelope, and idempotency keys for action submission, ensuring reliable and secure communication between AI agents and WebSocket servers.
Browse and install thousands of AI Agent skills in the Killer-Skills directory. Supports Claude Code, Windsurf, Cursor, and more.
ws-protocol is a technical specification outlining the protocol version, message envelope, and idempotency keys for action submission, ensuring reliable and secure communication between AI agents and WebSocket servers.
Writing Hookify Rules is a technique for creating markdown files with YAML frontmatter to define patterns and messages for AI agents.
AI Agentic API Testing Platform - Automated testing with specialized ephemeral agents
A collection of AI agent skills following the Agent Skills open format
moai-foundation-specs is a skill that provides expert-level SPEC lifecycle management, including version control, approval workflows, and integration with MoAI-ADK development pipelines.
supabase-audit-tables-list is a critical AI agent skill that ensures progressive file updates and logging for secure audit table listing in Supabase environments.
Evaluation is a process for assessing agent systems, requiring approaches that account for dynamic decision-making and non-deterministic behavior.
ralph-loop is an autonomous task loop that keeps Claude working on a task until it reaches a defined completion state.
chatgpt-bridge is a skill that generates machine-readable JSON output, following specific rules and protocols for efficient data exchange and processing.
prompt-optimizer is a skill that uses ultrathink principles to optimize and refine prompts, making them inevitable and effective for AI agents and applications.
semantik-plugin-development is a skill that allows developers to create plugins for Semantik, a self-hosted semantic search engine, to extend its capabilities for document ingestion, embedding, and other tasks.
full_auth is a skill that allows AI agents to perform phased implementation plans, executing all phases in sequence until completion, with explicit user permission.