Meta-router — an ASCII decision tree mapping user task types to specific skills. Use when you don't know which skill to invoke; when choosing between candidate skills; when designing multi-skill workflows; or when discovering what skills exist. Handles…
Use when writing a new skill for an AI agent (Claude Code, Copilot CLI, Cursor, Codex, Gemini CLI); when editing or improving an existing skill; when a skill produces inconsistent results; when designing the invocation pattern for a skill; when setting token…
Manage token budgets, progressive disclosure, context window optimization, summarization strategies, dual-representation compilation (human-readable vs agent-optimized), structured context pruning, attention budget allocation, context retention policies…
Use when designing or debugging AI agent context strategies, optimizing token budgets, building context assembly pipelines, or diagnosing context-related failures (wrong answers from missing info, bloated context causing poor reasoning, conversation drift).…
Use when minimizing the token cost of a given context payload while holding answer quality constant — the autonomous budget-saving engine. Handles context measurement and token accounting per level, minimization levers (minification, deduplication,…
Use when designing RAG pipelines, engineering prompts at scale, building LLM evaluation frameworks, implementing function calling and tool use, or optimizing LLM latency and cost. Handles RAG pipeline design (chunking strategies, embedding models, vector…
Use when minimizing LLM token consumption and cost, designing token budgets, optimizing prompt caching, compressing context, controlling output tokens, or measuring cost per request. Handles input/output token budget math, prompt-cache prefix economics,…
Use when building cross-platform mobile apps with Flutter, writing Dart code, designing widget trees, choosing state management (Riverpod, Bloc, Provider), implementing platform channels for native features, optimizing Flutter performance (Impeller/Skia,…