Apply whenever writing or editing source code in any core's build agent (backend-eng, ui-builder, or equivalent). Governs when a comment belongs at all and what it may say. Default is no comment; a comment is only for the non-obvious. Never restates what the code does, never narrates the task, fix, or conversation that produced it.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
Well-named identifiers and clear structure explain what code does. A comment earns its place only when it explains something the code cannot: a hidden constraint, a non-obvious invariant, a workaround for a specific external bug, or behavior that would surprise a reader.
Most lines, functions, and blocks need zero comments. Before writing one, check whether the same clarity is reachable by renaming a variable or function instead — prefer that over a comment every time.
Only for the non-obvious:
counter++).When editing a file that already has comments violating these rules, remove or rewrite them as part of the same change rather than leaving them in place — don't let a nearby edit normalize the pattern.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Push the LLM to reconsider, refine, and improve its recent output. Use when user asks for deeper critique or mentions a known deeper critique method, e.g. socratic, first principles, pre-mortem, red team
The English summary is being prepared. Showing the original description.
Facilitate a brainstorming session using diverse creative techniques. Use when the user says 'help me brainstorm' or 'help me ideate'
The English summary is being prepared. Showing the original description.
Research a topic to support a decision, three ways: draft a research prompt for the user to run in their own tool (ChatGPT, Gemini, Grok, Perplexity, …), turn a finished research report into a short summary with cited sources that other skills can use directly, or run the research here with parallel web searches. Built-in research types: market, domain, technical, competitive, user-voice, academic-lit; also supports choosing between candidates, and custom types via overrides. Use when the user says "deep recon", "research this", "draft a research prompt", "process this research report", "market research", "domain research", "technical research", "competitor research", "literature review", or "help me choose between"
The English summary is being prepared. Showing the original description.
Test a half-formed idea in a questioning conversation, with different personas probing its weak points, until the user can act on it or drop it with confidence. Optionally writes a short brief for planning skills to build on. Use when the user says 'forge an idea', 'pressure-test this idea', 'stress-test my thinking', or 'harden this idea'
The English summary is being prepared. Showing the original description.
Create, update, or validate a PRD. Use when the user wants help producing, editing, or validating a PRD
The English summary is being prepared. Showing the original description.
Test a product concept with Amazon's Working Backwards method: write the press release for the finished product first, then answer hard customer and stakeholder questions, ending in a complete PRFAQ document. Use when the user requests to 'create a PRFAQ', 'work backwards', or 'run the PRFAQ challenge'
The English summary is being prepared. Showing the original description.