designing hybrid context layers
by loreto
Architects the right retrieval strategy for every query — teaching your agent when to use RAG, a knowledge graph, or a temporal index instead of defaulting to vector search for everything.
Skills for retrieval pipelines, evaluations, embeddings, MCP servers, and running language models in production. Measure and operate your AI stack instead of guessing.
1,003 skills · page 1 of 42
by loreto
Architects the right retrieval strategy for every query — teaching your agent when to use RAG, a knowledge graph, or a temporal index instead of defaulting to vector search for everything.
by loreto
RAG fails quietly. It retrieves documents, returns confident-looking answers, and misses the question entirely — because the question required connecting facts across documents, reasoning about sequence, or tracing causation. This skill gives you a five-question diagnostic checklist that classifies any failing query as either RAG-safe or structurally RAG-incompatible, then maps it to the specific failure pattern and the architectural fix that resolves it.
by tudor
Build a full-stack AI chatbot trained on your own documents across any industry — legal, healthcare, e-commerce, HR, finance, real estate, insurance, education, cybersecurity, government, and more.
by Roy Yuen
Professional prompt engineering patterns for building robust, secure, and production-ready LLM applications.
by Roy Yuen
Design, debug, and optimize production RAG systems with expert architecture, hybrid search, and grounding strategies.
by Arun Dass
Cuts AI coding agent token consumption by 75–94%
Drastically reduce RAG costs and latency while improving retrieval accuracy through advanced memory architecture.
Independently audit AI outputs for accuracy, evidence, instruction compliance, and release readiness.
Let your AI agent design real Xfer Serum 2 presets and themed preset banks from plain-language requests.
Scaffold a secure, spec-compliant MCP server from a description of the tools you want to expose. Sets up the official SDK (TypeScript or Python/FastMCP), defines tools/resources/prompts with strict JSON Schema, wires the right transport (stdio or Streamable HTTP), adds OAuth 2.1 for remote, and hardens against the MCP-specific footguns — prompt injection via tool output, token passthrough, over-broad scopes, command/path/SSRF injection, leaked secrets — before it ships. Returns a runnable skeleton plus a security checklist. Built by someone who's shipped production MCP servers.
by Etka Ozer
Equip your agent with a persistent, local Markdown-based "second brain" for long-term project memory.
Turn sourced model-release or retirement notices into a scoped review queue: affected workflows, excluded deployment routes, unknowns and deadlines.
Produces a diagnosis of why a RAG system gives confident-but-wrong answers, or picks between vector search, knowledge graph, and structured/temporal retrieval. Use when the user says "why is our RAG hallucinating", "diagnose this failing query", "should we use a knowledge graph", "pick a retrieval architecture", or "design our memory layer". Do not use when the request is building a RAG system from zero.
Design and write the eval suite for your LLM-powered feature — the metrics that match your failure modes, a golden dataset plan with starter cases, anchored rubrics, LLM-as-judge prompts with the known bias mitigations, and pass/fail gates wired for CI.
by Edric Vale
Drive a running TouchDesigner instance through the twozero MCP bridge: create operators, wire nodes, animate parameters, run render passes, and query state back
by Ignasi
Designs, builds, tests and connects production-ready MCP servers in TypeScript or Python. Task-oriented tools with LLM-friendly descriptions, strict schemas, safe auth, token-cheap outputs, Inspector + agent evals, and configs for Claude Code, Desktop, Cursor and VS Code.
by Ignasi
Builds evaluation datasets for prompts, chatbots, RAG and agents: coverage matrix, realistic, edge, adversarial and refusal cases, deterministic graders and LLM-judge rubrics, validated JSONL (duplicates, leaks, schema), frozen test split, harness config and baseline report.
by Kaymue
Diagnose broken RAG systems. 8 failure categories: chunking, embeddings, retrieval, reranking, hallucination. Recall@k measurement.
by Hermes Test
Defensive vulnerability audit for PHP and COBOL stacks using live CISA KEV and CVE threat data.
by Edric Vale
Let your agent drive real web pages via accessibility snapshots and stable element handles — fewer tokens and no fragile coordinate guessing.
by Edric Vale
Use to drive Blender from an agent over a socket: install the blender-mcp addon, connect, create primitives, assign materials, run operations, render images, an
by RuH
Generate complete, standards-compliant favicon sets and PWA icons from a single logo file.
Full Apple Music playback, catalog discovery, Spotify import & autonomous harmonic DJ transitions for macOS
by GTDataworks
Capture tool-trace receipts, label failures, and emit replay prompts so you can debug an agent run after it finishes.