Jsnnmsc
UserFocused tools for software engineers to size up a change before making it: scope, impact, tradeoffs, decision points, and a first-principles pass on your own read of the problem — plus cheap subagents for the mechanical work. In your language.
Categories
Indexed Skills (11)
3w
Break a possible adjustment or change down into Where, What, Why — where it applies, what actually changes, why it matters — plus what to do about it. Trigger when the user describes a possible tweak/adjustment and wants a quick breakdown, or says "3w", "where what why", "give me the 3w". Output: Where + What + Why + bottom line. Max 10 lines.
catchup
Quick status brief so the user can pick up where things left off — what changed, what state it's in, what's next. Trigger when the user says "catch up", "catch me up", "where were we", "what's the status", "what's going on", or returns to a task/repo after a gap and wants an overview. Output: since last time / current state / next step. Max 12 lines. Grounded in actual git status/log/diff, not guesses.
decision
Surface the actual decision points in a task, plan, or problem — the specific choices that need to be made, the options for each, and a default recommendation. Trigger when the user wants decision points before diving in, or says "decision points", "what do I need to decide", "give me the decision points". Output: numbered decision points, each with options and a default + bottom line. Max 10 lines.
impact
Surface the blast radius of a change — which modules, services, users, or data get touched, directly or downstream — plus what to do about it. Trigger when the user asks "what does this affect", "blast radius", "who gets touched by this", "impact of this change", or pastes a change and wants to know what depends on it. Output: directly affected + downstream affected + who notices + bottom line. Max 10 lines.
tldr
Quick plain-language brief — skip the explanation, get the point. Use when the user says "tldr", "quick version", "what is X", "should I use X", "just tell me", "give me the short version", or is clearly in a hurry. Different from one-line: one-line gives one sentence and stops. tldr adds the two facts that matter and a bottom line. Output: bold topic + one sentence + 2 adaptive bullets + bottom line. Max 5 lines. No diagrams, no analogies, no history.
clear-view
Explain the current situation as one short plain-language paragraph — what we thought, what turned out to be true, where that leaves us — then the one decision to make. Trigger when the user says "clear view", "so where are we", "what's the situation", "I don't follow the current state", "what do I need to decide", or after a long investigation that needs to converge on one decision point. Different from tldr: tldr explains a thing in 5 clipped lines. clear-view tells the story of the situation in plain prose, with the context the user is missing, and ends on a decision. Output: one narrative paragraph + one bold decision line. No bullets, no headers. Max 8 lines.
first-principle
Strip a problem down to what's actually true underneath — separate inherited convention from real constraint, then restate the problem as it really is. Trigger when the user says "first principle", "第一性原理", "本質是什麼", "從頭想一次", "從根本分析", "這問題的根源", "我們是不是搞錯問題了", "這個設計合理嗎", "why are we doing it this way", or is about to solve a problem the way it has always been solved. Different from tradeoff and decision: those work inside the problem as stated. This one checks whether the problem is stated right, and hands back a different one if it isn't. Output: the stated problem, the assumptions under it split into convention vs unverified, the facts that survive, and one line naming the real problem. Max 12 lines.
tradeoff
Quick A-vs-B comparison for a technical choice — what each option costs, what it buys, and which one to pick. Trigger when the user asks "X vs Y", "should I use X or Y", "which is better here", "what's the tradeoff between X and Y". Output: gain/cost for each option + pick. Max 10 lines.
doubt
Challenge whether a requirement, feature, or design is needed at all — before any effort goes into how to build it. Trigger when the user says "doubt", "do we actually need this", "is this necessary", "push back on this", "am I overbuilding", or is drowning in options and wants the list cut down instead of expanded. Different from tradeoff and decision: those help you pick between options, this one questions whether the thing should exist and kills options rather than adding them. Output: the claimed problem, a plain-language paragraph on whether it's real, then do-nothing / smallest-version / one verdict. Max 10 lines.
scope
Turn a vague feature request or task description into a clear scope — what's in, what's out, what's unknown, and how big it is. Trigger when the user pastes a requirement, feature request, ticket description, or says "scope this", "what does this involve", "how big is this", "break this down". Output: in scope / out of scope / unknowns / complexity estimate. Max 15 lines.
one-line
Explain it in exactly one sentence — nothing else. Trigger when the user says "one line", "one sentence", "in a sentence", "一句話", "explain in one line", or asks what something is and clearly wants no more than that. Different from tldr: tldr gets 5 lines and can add key facts or a verdict. one-line gets one sentence and stops. Output: one bold sentence. No bullets, no heading, no follow-up offer.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.