Maintain a one-page burn-down spec as the working memory of a large or multi-session task. Use when work spans sessions, when the user asks for a spec or brief, or when a grill ends on a task too big to carry in a confirmed summary alone.
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
Intent: one markdown file holding the task's remaining intent, so any fresh session re-bootstraps by reading it. It is working memory, not documentation.
Shape (suggested, not enforced), hard cap ~1 page:
No file-by-file implementation plan — that lives in the working session and dies with it.
Guardrails:
specs/ (gitignored, dies with the task,
never lands on main). A spec big enough that its burn-down history or surviving
slivers have long-term value lives in wiki/specs/ instead — committed, and listed
in the wiki index under a separate "Specs (working memory)" section so queries never
mistake burn-down state for settled knowledge.Exit: when the acceptance criteria verify, distill anything worth keeping into the wiki (the llm-wiki skill decides what and in what shape), then delete the spec file — git keeps the history; a wiki-resident sliver may survive as (or merge into) a durable page. A leftover spec is a visible signal of an unfinished task.
还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Records hypotheses, evidence and decisions when proposing, running or evaluating forecast experiments, including backtests and feature or model comparisons. Excludes routine tests and unchanged forecast operation.
中文简介正在准备中。当前显示原文说明。
Diagnose GitHub CLI hiccups caused by Codex sandbox constraints; not a general GitHub workflow.
中文简介正在准备中。当前显示原文说明。
Review a design artifact (spec, plan, PR packet) with fresh, memoryless subagents. Invoke when the user asks — "goldfish this", "fresh eyes", "tear this apart" — or after suggesting it for a spec that has grown load-bearing and the user agrees.
中文简介正在准备中。当前显示原文说明。
Interrogate the user before implementing a non-trivial or ambiguous ask, converging on shared intent. Use when the user says "grill me", when an ask admits multiple interpretations, or before any sizable piece of work whose intent isn't fully clear.
中文简介正在准备中。当前显示原文说明。
Launches and supervises long-running commands — solves, workflows, downloads, builds — detached with a log, PID and exit sentinel, then tells stuck from slow and kills verifiably. Use before running anything that may exceed a minute, for nohup or background launches, and when a run hangs, shows no output, or timed out.
中文简介正在准备中。当前显示原文说明。
Build and maintain a persistent, LLM-maintained markdown wiki following Andrej Karpathy's LLM Wiki pattern — the agent incrementally compiles raw sources into an interlinked knowledge base instead of re-deriving answers from scratch. Use this whenever the user wants to ingest a source (an article, paper, transcript, chapter, meeting notes), file knowledge or conversation findings into the wiki, query the accumulated knowledge base, lint or health-check the wiki, or mentions "the wiki", "knowledge base", "ingest this", "file this", "what do we know about X", or accumulating notes on a topic over time — even if they don't say "LLM Wiki" explicitly.
中文简介正在准备中。当前显示原文说明。