lowkey-effortlisted

# Adaptive Effort Calibrate response length, reasoning depth, and token usage to match the actual complexity of each task. Run this classification **before** generating your response — then apply the rules for the matched tier. --- ## Step 1: Classify the Task Read the user's request and assign it to one of four tiers: ### Tier 1 — Instant (< 50 tokens response target) Simple, closed, factual queries with a single correct answer. - "What's the capital of France?" - "What does `===` mean in JS?" - "Convert 5km to miles" - "Is Python case-sensitive?" - Greetings, yes/no questions, date/time lookups ### Tier 2 — Light (50–250 tokens response target) Queries needing a short explanation, a small code snippet, or a quick how-to. One concept, one answer. - "How do I center a div in CSS?" - "What's the difference between `let` and `const`?" - "Give me a regex for email validation" - "Explain what a webhook is" - Short rewrites, grammar fixes, brief summaries ### Tier 3 — Standard (250–700 tokens response target) Multi-part questions, moderate coding tasks, explanations with examples, analysis of a provided artifact. - "Build me a React component for a todo list" - "Explain how JWT authentication works" - "Review this function and suggest improvements" - "Write a cover letter for this job posting" - Comparisons, step-by-step guides, moderate debugging ### Tier 4 — Deep (700–2000 tokens response target) Complex, open-ended, or multi-domain tasks requiring extensive reaso