recap
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinese/recap — Session Orientation & Awareness
/recap — 会话定位与感知
Goal: Orient yourself fast. Rich context by default. Mid-session awareness with .
--now目标:快速了解当前定位。默认提供丰富上下文。使用参数可获取会话中途的感知信息。
--nowUsage
使用方法
/recap # Rich: retro summary, handoff, tracks, git
/recap --quick # Minimal: git + focus only, no file reads
/recap --now # Mid-session: timeline + jumps from AI memory
/recap --now deep # Mid-session: + handoff + tracks + connections/recap # 丰富模式:回顾总结、工作交接、任务跟踪、Git状态
/recap --quick # 极简模式:仅显示Git状态和专注度,不读取文件
/recap --now # 会话中途模式:AI记忆中的时间线与跳转记录
/recap --now deep # 会话中途深度模式:+ 工作交接 + 任务跟踪 + 关联信息DEFAULT MODE (Rich)
默认模式(丰富版)
Run the rich script, then add suggestions:
bash
bun ~/.claude/skills/recap/recap-rich.tsScript reads retro summaries, handoff content, tracks, git state. Then LLM adds:
- What's next? (2-3 options based on context)
执行丰富版脚本,然后添加建议:
bash
bun ~/.claude/skills/recap/recap-rich.ts脚本会读取回顾总结、工作交接内容、任务跟踪记录、Git状态。之后LLM会补充:
- 下一步建议?(基于上下文提供2-3个选项)
Step 2: Git context
步骤2:Git上下文
bash
git status --short
git log --oneline -1Check what's appropriate from git status:
- Uncommitted changes? → show them, suggest commit or stash
- On a branch (not main)? → to see branch work
git log main..HEAD --oneline - Branch ahead of remote? → suggest push or PR
- Clean on main? → just show last commit, move on
Only read what matters — don't dump 10 commits if status is clean.
bash
git status --short
git log --oneline -1根据Git状态显示合适的信息:
- 存在未提交的更改? → 显示更改内容,建议提交或暂存
- 处于非main分支? → 执行查看分支上的工作内容
git log main..HEAD --oneline - 分支领先于远程仓库? → 建议推送或创建PR
- main分支状态干净? → 仅显示最后一次提交记录,继续后续流程
只读取关键信息——如果状态干净,不要输出10条提交记录。
Step 3: Read latest ψ/ brain files
步骤3:读取最新的ψ/ 脑图文件
Sort all ψ/ files by modification time, read the most recent:
bash
find ψ/ -name '*.md' -not -name 'CLAUDE.md' -not -name 'README.md' -not -name '.gitkeep' 2>/dev/null | xargs ls -t 2>/dev/null | head -5Read those top 5 files. This recovers the same context restores — handoffs, retros, learnings, drafts, whatever was touched last.
/compact按修改时间排序所有ψ/目录下的文件,读取最新的5个:
bash
find ψ/ -name '*.md' -not -name 'CLAUDE.md' -not -name 'README.md' -not -name '.gitkeep' 2>/dev/null | xargs ls -t 2>/dev/null | head -5读取这5个文件。这能恢复与/compact命令相同的上下文——工作交接、回顾总结、学习笔记、草稿等最近操作过的内容。
Step 4: Dig last session
步骤4:挖掘上一次会话
bash
ENCODED_PWD=$(pwd | sed 's|^/|-|; s|/|-|g')
PROJECT_BASE=$(ls -d "$HOME/.claude/projects/${ENCODED_PWD}" 2>/dev/null | head -1)
export PROJECT_DIRS="$PROJECT_BASE"
python3 ~/.claude/skills/dig/scripts/dig.py 1Include in recap:
📡 Last session: HH:MM–HH:MM (Xm, N msgs) — [topic]Need more? or .
/dig 5/dig --timelineTotal: 1 bash call + LLM analysis
bash
ENCODED_PWD=$(pwd | sed 's|^/|-|; s|/|-|g')
PROJECT_BASE=$(ls -d "$HOME/.claude/projects/${ENCODED_PWD}" 2>/dev/null | head -1)
export PROJECT_DIRS="$PROJECT_BASE"
python3 ~/.claude/skills/dig/scripts/dig.py 1在回顾中包含以下内容:
📡 上一次会话:HH:MM–HH:MM(时长X分钟,N条消息)—— [主题]需要更多内容?使用或。
/dig 5/dig --timeline总计:1次bash调用 + LLM分析
QUICK MODE (/recap --quick
)
/recap --quick极简模式(/recap --quick
)
/recap --quickMinimal, no content reads:
bash
bun ~/.claude/skills/recap/recap.tsScript outputs git status + focus state (~0.1s). Then LLM adds:
- What's next? (2-3 options based on git state)
极简模式,不读取内容:
bash
bun ~/.claude/skills/recap/recap.ts脚本输出Git状态 + 专注度状态(约0.1秒)。之后LLM会补充:
- 下一步建议?(基于Git状态提供2-3个选项)
"What's next?" Rules
“下一步建议?”规则
| If you see... | Suggest... |
|---|---|
| Handoff exists | Continue from handoff |
| Untracked files | Commit them |
| Focus = completed | Pick from tracks or start fresh |
| Branch ahead | Push or create PR |
| Streak active | Keep momentum going |
| 当你看到... | 建议... |
|---|---|
| 存在工作交接记录 | 从工作交接的内容继续 |
| 存在未跟踪文件 | 提交这些文件 |
| 专注度状态 = 已完成 | 从任务跟踪中选择或开始新任务 |
| 分支领先于远程 | 推送或创建PR |
| 处于活跃工作流 | 保持当前节奏 |
Hard Rules
硬性规则
- ONE bash call — never multiple parallel calls (adds latency)
- No subagents — everything in main agent
- Ask, don't suggest — "What next?" not "You should..."
- 仅一次bash调用 — 绝不进行多并行调用(会增加延迟)
- 不使用子Agent — 所有操作都在主Agent中完成
- 询问而非命令 — 用“下一步做什么?”而非“你应该...”
NOW MODE (/recap --now
)
/recap --nowNOW模式(/recap --now
)
/recap --nowMid-session awareness from AI memory — no file reading needed. Use when user asks "where are we", "now", "status", "what are we doing".
AI reconstructs session timeline from conversation memory:
markdown
undefined基于AI记忆的会话中途感知 — 无需读取文件。适用于用户询问“我们到哪一步了”“现在情况”“状态”“我们正在做什么”等场景。
AI会从对话记忆中重建会话时间线:
markdown
undefinedThis Session
当前会话
| Time | Duration | Topic | Jump |
|---|---|---|---|
| HH:MM | ~Xm | First topic | - |
| HH:MM | ~Xm | Second topic | spark |
| HH:MM | ongoing | Now: Current | complete |
Noticed:
- [Pattern - energy/mode]
- [Jump pattern: sparks vs escapes vs completions]
Status:
- Energy: [level]
- Loose ends: [unfinished]
- Parked: [topics we'll return to]
My Read: [1-2 sentences]
Next?
undefined| 时间 | 时长 | 主题 | 跳转类型 |
|---|---|---|---|
| HH:MM | ~X分钟 | 第一个主题 | - |
| HH:MM | ~X分钟 | 第二个主题 | spark |
| HH:MM | 进行中 | 当前:当前任务 | complete |
观察到:
- [模式 - 精力/状态]
- [跳转模式:sparks vs escapes vs completes]
状态:
- 精力:[等级]
- 未完成事项:[未完成内容]
- 已暂停事项:[后续将回归的主题]
我的解读:[1-2句话]
下一步?
undefinedJump Types
跳转类型
| Icon | Type | Meaning |
|---|---|---|
| spark | New idea, exciting | |
| complete | Finished, moving on | |
| return | Coming back to parked | |
| park | Intentional pause | |
| escape | Avoiding difficulty |
Healthy session: Mostly sparks and completes
Warning sign: Too many escapes = avoidance pattern
| 图标 | 类型 | 含义 |
|---|---|---|
| spark | 新想法,令人兴奋 | |
| complete | 已完成,继续推进 | |
| return | 回归已暂停的事项 | |
| park | 有意暂停 | |
| escape | 回避困难 |
健康的会话:主要包含sparks和completes类型
警告信号:过多的escapes类型 = 回避模式
NOW DEEP MODE (/recap --now deep
)
/recap --now deepNOW深度模式(/recap --now deep
)
/recap --now deepSame as but adds bigger picture context.
--now与模式相同,但增加全局上下文信息。
--nowStep 1: Gather (parallel)
步骤1:收集信息(并行)
1. Current session from AI memory
2. Read latest handoff: ls -t ψ/inbox/handoff/*.md | head -1
3. Git status: git status --short
4. Tracks: cat ψ/inbox/tracks/INDEX.md 2>/dev/null1. AI记忆中的当前会话记录
2. 读取最新的工作交接记录:ls -t ψ/inbox/handoff/*.md | head -1
3. Git状态:git status --short
4. 任务跟踪:cat ψ/inbox/tracks/INDEX.md 2>/dev/nullStep 2: Output
步骤2:输出内容
Everything from , plus:
--nowmarkdown
undefined包含模式的所有内容,再加上:
--nowmarkdown
undefinedBigger Picture
全局视角
Came from: [Last session/handoff summary - 1 line]
Working on: [Current thread/goal]
Thread: [Larger pattern this connects to]
来源:[上一次会话/工作交接总结 - 1行]
当前工作:[当前任务线程/目标]
任务关联:[当前任务所属的更大模式]
Pending
待办事项
| Priority | Item | Source |
|---|---|---|
| Now | [Current task] | This session |
| Soon | [Next up] | Tracks/discussion |
| Later | [Backlog] | GitHub/tracks |
| 优先级 | 事项 | 来源 |
|---|---|---|
| 立即 | [当前任务] | 本次会话 |
| 近期 | [下一步任务] | 任务跟踪/讨论记录 |
| 远期 | [待办清单] | GitHub/任务跟踪 |
Connections
关联信息
Pattern: [What pattern emerged]
Learning: [Key insight from session]
Oracle: [Related past pattern, if any]
My Read: [2-3 sentences - deeper reflection]
Next action?
---模式:[出现的模式]
学习收获:[会话中的关键见解]
参考过往:[相关的过往模式(如有)]
我的深度解读:[2-3句话 - 深度反思]
下一步行动?
---Session Context
会话上下文
The recap scripts ( and ) auto-detect and display the current session:
recap.tsrecap-rich.ts📡 Session: 74c32f34 | arra-oracle-skills-cli | 2h 15mDetection: scans for the most recent session file, extracts short ID and elapsed time from first timestamp.
~/.claude/projects/[encoded-pwd]/*.jsonlIf session detection fails, skip silently — it's informational only.
回顾脚本(和)会自动检测并显示当前会话:
recap.tsrecap-rich.ts📡 会话:74c32f34 | arra-oracle-skills-cli | 2小时15分钟检测方式:扫描目录下最新的会话文件,从第一个时间戳中提取短ID和已耗时。
~/.claude/projects/[encoded-pwd]/*.jsonl如果会话检测失败,将静默跳过——这部分仅为信息展示。
Demographics Context
用户画像上下文
If CLAUDE.md contains demographics from wizard v2, include in recap output:
/awakenmarkdown
**Oracle**: [name] ([pronouns]) | **Human**: [name] ([pronouns]) | **Language**: [pref]Add this as one line after the timestamp in any mode. If demographics not present, skip silently.
Look for fields in CLAUDE.md: , , , , .
Human PronounsOracle PronounsLanguageTeamExperiencePhilosophy: Detect reality. Surface blockers. Offer direction. "Not just the clock. The map."
Version: 8.0 (Merged where-we-are into --now mode)
Updated: 2026-02-10
如果CLAUDE.md文件中包含来自向导v2的用户画像信息,需在回顾输出中包含:
/awakenmarkdown
**Oracle**:[名称]([代词]) | **人类用户**:[名称]([代词]) | **语言**:[偏好]在任何模式下,将此内容添加到时间戳之后的一行。如果没有用户画像信息,则静默跳过。
在CLAUDE.md中查找以下字段:、、、、。
Human PronounsOracle PronounsLanguageTeamExperience理念:洞察实际情况。明确阻碍因素。提供行动方向。“不只是时钟,更是地图。”
版本:8.0(将where-we-are功能合并到--now模式中)
更新日期:2026-02-10