my-favorite-prompts

Invoke immediately if the Navigator mentions: inspecting changes (status, diff, log, show), staging/adding files (add, stage), creating commits (commit, amend, reset), branch management (branch, checkout, merge, rebase), remote operations (push, pull, fetch), or repository state (stash, worktree). Trigger words: git, commit, status, diff, add, stage, branch, merge, push, pull, log, revert, stash, worktree, 'what changed', 'save changes', 'commit this', 'show me the diff'.

Use when creating, updating, troubleshooting, or organizing Ubuntu GNOME dock buttons, .desktop launchers, Zenity control menus, script shortcuts, launcher icons, or taskbar actions for commands stored across multiple projects.

Use when investigating opencode sessions, subsessions, agent drift, missed handoffs, tool misuse, or feedback grounded in opencode session exports.

Guide a user with basic CSS knowledge who is learning Tailwind CSS and frontend styling. Use when the user asks for help styling a page, layout, component, tag, or UI state and wants coaching through style categories, good frontend practices, CSS mental models, or Tailwind decision-making without being given the finished code solution.

Use when investigating swap, memory pressure, compressor usage, llama.cpp memory, Metal/unified memory, or model-server RAM issues on the Mac Studio ai-server.

Run any question, idea, or decision through a council of 5 AI advisors who independently analyze it, peer-review each other anonymously, and synthesize a final verdict. Based on Karpathy's LLM Council methodology. MANDATORY TRIGGERS: 'council this', 'run the council', 'war room this', 'pressure-test this', 'stress-test this', 'debate this'. STRONG TRIGGERS (use when combined with a real decision or tradeoff): 'should I X or Y', 'which option', 'what would you do', 'is this the right move', 'validate this', 'get multiple perspectives', 'I can't decide', 'I'm torn between'. Do NOT trigger on simple yes/no questions, factual lookups, or casual 'should I' without a meaningful tradeoff (e.g. 'should I use markdown' is not a council question). DO trigger when the user presents a genuine decision with stakes, multiple options, and context that suggests they want it pressure-tested from multiple angles.

Use when writing new code or reviewing existing code to decide which comments to add, keep, or remove. Defines what makes a comment valuable vs. noise.

Use when work involves LLM chat templates, tokenizer artifacts, special/stop/control tokens, reasoning tags, generation prompts, generation service task docs, local generation models, ONNX causal-LM contracts, SmolLM/Qwen/Gemma prompts, KV cache inputs, past_key_values, or thinking-mode sessions.

Use when you need to rename an opencode session - bypasses the interactive /rename command by updating the SQLite database directly

Use when extracting technical details from model artifacts, tokenizer configs, chat templates, special/stop/control tokens, ONNX causal-LM contracts, prompt files, generation service task docs, local generation models, SmolLM/Qwen/Gemma artifacts, KV cache inputs, past_key_values, or session exports where raw control text could break or steer an agent.

Use before creating, modifying, or reviewing code with singleton state, memoization, lazy loading, caches, registries, mutexes, unload/reset paths, or long-lived external resources

Use when a user asks to visualize task, subtask, roadmap, checklist, or project-plan dependencies in a terminal-readable format, especially for deciding worktrees, agent parallelization, vertical ASCII graphs, or what can run in parallel.

Use before creating, modifying, reviewing, or refactoring spec test files. Better Specs compliance, especially around one behavior per example, meaningful example names, describe/context structure, behavior-over-implementation assertions, `expect` syntax, or spec readability and maintainability problems.