33 awrshift

tour Skill

使用用户的真实项目文件,交互式演示 Memory Kit 系统。当用户说 "/memory-kit:tour"、"give me a tour"、"покажи как это работает" 时,或紧跟 /memory-kit:setup 之后使用。

安装方式:把技能目录放入 ~/.claude/skills/(Claude Code)或在 claude.ai 设置中启用;也可复制右侧安装命令一键添加。

查看源码

技能指令原文(SKILL.md)

/memory-kit:tour — Interactive walkthrough

You are giving the user a guided tour of their Claude Memory Kit. Use their actual files (not generic descriptions) so they see how the abstraction maps to concrete files.

Tour structure (10-15 minutes)

Stop 1 — the working agreement (injected, not a file in this repo)

Run /context or point at the session's opening block. Say: "The rules I work by ship with the plugin and are injected at every session start — there's nothing for you to maintain. Your own project rules, if you ever want any, live in .claude/rules/."

Stop 2 — Session entry (handoff + backlog)

Open the newest context/handoffs/*.md (or the template if none yet) and, if multi-project mode, projects//BACKLOG.md. Say: "This is the note the previous session left for me — the handoff. The start-hook injects the newest one automatically, so I open every session already knowing where we left off."

Stop 3 — Hot path (MEMORY.md)

Open .claude/memory/MEMORY.md. Say: "This is my hot cache. Date-tagged one-liners of patterns from recent sessions; the header on top is the current state of work, replaced at every close. A hook watches three size caps so this file physically can't silently bloat."

Stop 4 — Rules (.claude/rules/)

List any .claude/rules/*.md files. If folder is empty (only _example.md.disabled), say: "This is where hard project rules live. 'Don't use X', 'always check Y'. A rule loads every session, or — when it carries a paths: glob — only once a matching file is read. Empty for now — they'll appear when you start dictating rules and I propose them on /memory-kit:close-session."

Stop 5 — Knowledge concepts

List knowledge/concepts/*.md. Say: "Deep memory with facts. 'What's our typography scale', 'what we know about SEO for AI'. Reference articles. Empty now — I write them during /memory-kit:close-session when a pattern has repeated on three-plus dates and you say yes."

Stop 6 — Handoffs

Show context/handoffs/. Say: "One note per closed session; the newest one opens the next session. An old note nothing points at is deleted — git keeps the past, so 'what did we do last Tuesday?' still has an answer."

Stop 7 — Projects

List projects/*/ and open one README.md. Say: "Each client or product gets a folder, and its README is the map — where this project's tasks, specs, research, decisions and QA live. Memory is shared across all of them; paperwork is per project. Say 'we're working on ' and I switch to that one."

Stop 8 — Hooks

Don't deep-dive — the hooks ship inside the plugin, there is nothing in this repo to maintain. Say: "The hooks in the plugin's hooks/hooks.json run silently: one injects your memory and the working rules at every session start, one holds a manual compaction until state is saved, one asks before an existing test gets edited, guards stop destructive git commands, catch a zsh command that would split a variable into words differently than bash, and keep secret files like .env out of the conversation, and one logs the close."

Stop 9 — Operators

List what this plugin adds — one operator per folder in the plugin's skills/: /memory-kit:close-session, /memory-kit:tour, /memory-kit:setup, /memory-kit:memory-audit, /memory-kit:system-audit, /memory-kit:code-sync, /memory-kit:document, /memory-kit:session-review, /memory-kit:second-opinion, /memory-kit:qa-sweep. Say: "/memory-kit:close-session is the one that matters — the end-of-session ritual where I audit what happened, propose what's worth remembering forever, and leave the handoff note for next time. memory-audit is the surgery when the cache outgrows its caps; system-audit is the periodic look at whether this whole system still earns its place. code-sync reconciles specs with the code after a merge and document writes a PR body or changelog from the diff; session-review, second-opinion and qa-sweep are for when you build with subagents — you don't need any of these to start."

Closing

Ask: "Anything you want to drill into deeper? Or should we start with a real task — name your first project?"

What you do NOT do

  • Don't read every file in full. Show the first 5-10 lines so user sees the format, not the content.
  • Don't lecture. This is a tour, not a manual. Each stop = 1-2 sentences.
  • Don't propose changes during the tour. This is informational only. If user asks "can we change X" during tour — note it for after.
  • Don't skip stops based on emptiness. Empty folders are part of the story — show them and explain when they fill.

Length

Target 10-15 minutes total. If user is engaged and asks questions, fine — go longer. If user seems impatient, compress to 5 minutes by collapsing stops 4-7 into one "and these are the on-trigger layers" sweep.

配合站内内容