Agent Skills Hub

Extract structured data from unstructured files (PDF, PPTX, DOCX...)

Implement LlamaExtract for robust structured data extraction from PDF, DOCX, and PPTX files using Pydantic schemas.

mcp-development

Guidance for Model Context Protocol (MCP) server development, including tool design, resource handling, and AI/ML integration patterns.

speckit-plan

Execute the implementation planning workflow, generate technical design artifacts, and structure research tasks for Spec Kit projects.

markdown-token-optimizer

Analyzes markdown files to identify token-wasting patterns, providing actionable suggestions to optimize documentation for LLM consumption and token efficiency.

trulens-evaluation-workflow

A systematic workflow to instrument, evaluate, and monitor LLM applications using TruLens, supporting frameworks like LangChain, LangGraph, and LlamaIndex.

semantic-compression

Aggressively prune grammatical scaffolding and filler text from inputs to optimize LLM token usage while retaining core semantic content.

self-improve

Self-maintaining skill for OpenCode agents to update documentation, capture learnings, and extend tool/agent capabilities dynamically.

rag-engineer

Architect and optimize production-grade RAG systems. Master embedding models, vector databases, chunking strategies, and retrieval pipelines for high-accuracy LLM applications.

test-delete-skill

Unified AI gateway for 100+ LLMs with OpenAI-compatible API, model fallbacks, load balancing, and enterprise-grade tools.

agent-architecture

Implement production-grade AI agents with LangGraph, tool-calling guardrails, SSE streaming, and episodic memory. Includes anti-patterns, fix pairs, and stateful architecture patterns.

rag-implementation

Build production-grade RAG systems using vector databases, semantic search, and LangGraph to ground LLMs in external knowledge.

prompt-estruturado

A structured prompting framework to transform casual inputs into professional, modular LLM prompts with persona, context, task, format, and guardrails.