What makes Session Branch notable?

1.1K downloads. Trust evidence available

How should Session Branch be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Session Branch

Branch a coding session into a new conversation with full context handoff — generate structured handoff doc and startup prompts. Triggers ONLY on '支线任务' / '开...

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.1k

Updated

Oct 11, 2026

Version

1.4.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.1K downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 1.1K downloadsadoption · observed Oct 11, 2026
Latest release: 1.4.0release · observed Jul 15, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s177q4wcvafq6fzfkhk2g3cwth83y01d:session-branch

Install using `clawhub skill install s177q4wcvafq6fzfkhk2g3cwth83y01d:session-branch` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/edwardwason/session-branch before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-edwardwason-session-branch/snapshot"

Documentation

CLAWHUB

144,882 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: "session-branch"
description: "Branch a coding session into a new conversation with full context handoff — generate structured handoff doc and startup prompts. Triggers ONLY on '支线任务' / '开个支线' / '分叉'. Do NOT use for new projects, general coding, or non-handoff session management."
version: "1.4.0"
slug: "session-branch"
displayName: "Session Branch"
summary: "将当前编码会话分叉到新对话，完整保留上下文。自动生成结构化交接文档和启动提示词。"
license: "MIT-0"
tags: ["session", "handoff", "branch", "context", "continuity"]
---

# Session Branch

Branch your current coding session into a new conversation without losing context.

## When to Use

- Current conversation is getting long and context compression is degrading quality
- You want to start a new task on the same project but keep full context
- You need to fork your work into a parallel direction
- User says: "支线任务" / "开个支线" / "分叉"

## Privacy & Consent

**Before executing any step, the agent MUST:**

1. **Warn the user** that this skill will write a file to the project directory (e.g., `docs/session-handoff.md`)
2. **Ask for explicit consent** before scanning any IDE-specific identity, memory, or configuration files
3. **Never harvest env var values** — only record variable names and whether they are configured (yes/no)
4. **Use project-relative paths** in all generated documents — never absolute paths containing usernames or home directories

## Execution Flow

### Step 1: Analyze Current Session

Scan the current conversation and project to extract:

1. **Primary request and intent** — all user intents, deduplicated, numbered chronologically
2. **Key technical concepts** — keywords, categorized (Architecture / Tools / Patterns)
3. **Files and code sections** — all Read/Edit/Write operations, with change summaries and key snippets
4. **Errors and fixes** — recent 5-10 errors with root cause and solution
5. **Problem solving** — completed achievements + key decision chain (chosen vs rejected + why)
6. **All user messages** — original text preserved **with sanitization** (strip credentials/PII), truncate >500 chars with "..."
7. **Conversation language** — detected primary language
8. **Current work** — active task, unfinished Todos with status, branchable directions

**Supplementary context** (not token-budgeted): project identity, platform status, env var names (status only), capability boundary, knowledge file index, user preferences.

#### IDE-Specific Additional Scanning

> **Consent required**: Before scanning any of the following, inform the user what files will be accessed and ask for explicit permission. Skip any category the user declines.

**For WorkBuddy**, also scan (with user consent):
- **Identity files**: `~/.workbuddy/SOUL.md`, `IDENTITY.md`, `USER.md` — persona and preferences
- **Memory files**: `.workbuddy/memory/MEMORY.md` + daily logs — project memory
- **Installed skills**: `~/.workbuddy/skills/` — list of active skills
- **Scheduled tasks**: automation/cron task list and status
- **Channel config**: I

README.md

# Session Branch — 支线任务

[![版本](https://img.shields.io/badge/版本-1.4.0-blue)](https://github.com/EdwardWason/session-branch)
[![许可证](https://img.shields.io/badge/许可证-MIT--0-green)](LICENSE)
[![ClawHub](https://img.shields.io/badge/ClawHub-session--branch-orange)](https://clawhub.ai/skills/session-branch)

将当前编码会话分叉到新对话，完整保留上下文。自动生成结构化交接文档、启动提示词，引导新会话无缝接续。

## 为什么需要

长对话中 LLM 上下文压缩会降低质量。你想开新对话，但不想丢失所有知识、决策和代码上下文。Session Branch 通过生成一份完整的交接文档，让新 AI 会话可以"热启动"你的项目。

## 功能

- **结构化交接文档** — 12 节模板，覆盖项目身份、决策链、数据流、能力边界、代码变更、平台状态等
- **验证检查清单** — 12 大类 40+ 项，确保不遗漏
- **IDE 专属启动提示词** — 预置 TRAE SOLO / WorkBuddy / Cursor / Claude Code 模板
- **三步启动流程** — 加载 → 汇报 → 问询，新 AI 先理解再行动
- **个人信息自动过滤** — 自动剥离真实姓名、路径、token 和密钥
- **可接续方向枚举** — 列出下一步可做什么，含前置条件和复杂度评估
- **WorkBuddy 深度适配** — 身份文件体系（SOUL.md/IDENTITY.md/USER.md）、记忆体系、技能列表、定时任务、IMA/飞书通道、MCP 连接器

## 快速开始

### 安装

```bash
clawhub install session-branch
```

### 使用

在当前对话中说：

- "支线任务" / "开个支线" / "分叉"

Skill 将：
1. 分析当前会话和项目
2. **写入文件**：在项目中生成 `docs/session-handoff.md`（如已存在会覆盖）
3. 给你一份新对话的启动提示词

> **注意**：此 Skill 会在项目目录中创建/覆盖文件。IDE 专属扫描（如 WorkBuddy 身份文件、TRAE memory 文件）需要你明确同意后才会执行。新会话中读取 memory/identity 文件是 OPT-IN 的，可以拒绝。交接文档中的用户消息会自动脱敏（删除凭证/PII/敏感信息）。

### 在新对话中

粘贴启动提示词，新 AI 将：
1. 读取交接文档 + 规则文件
2. 汇报它了解到的项目状态
3. 建议可接续的支线方向
4. 问你想走哪条路

## 配置

| 设置 | 默认值 | 说明 |
|------|--------|------|
| `handoff_path` | `docs/session-handoff.md` | 交接文档保存路径 |
| `include_checklist` | `true` | 是否用检查清单验证 |
| `target_ide` | `auto` | 目标 IDE（auto/trae/workbuddy/cursor/claude-code） |

## 文档导航

| 文档 | 说明 |
|------|------|
| [SKILL.md](SKILL.md) | Skill 规范与执行流程 |
| [交接文档模板](references/handoff-template.md) | 12 节交接文档完整模板 |
| [验证检查清单](references/checklist.md) | 12 大类 40+ 项验证清单 |
| [启动提示词模板](references/startup-prompts.md) | 各 IDE 专属启动提示词 |
| [贡献指南](docs/CONTRIBUTING.md) | 如何参与贡献 |
| [发布指南](docs/PUBLISHING_GUIDE.md) | 如何发布新版本 |
| [CHANGELOG.md](CHANGELOG.md) | 版本变更记录 |

## 许可证

MIT-0

---

# Session Branch

[![version](https://img.shields.io/badge/version-1.4.0-blue)](https://github.com/EdwardWason/session-branch)
[![License](https://img.shields.io/badge/license-MIT--0-green)](LICENSE)
[![ClawHub](https://img.shields.io/badge/ClawHub-session--branch-orange)](https://clawhub.ai/skills/session-branch)

Branch your current coding session into a new conversation with full context handoff. Generate a structured handoff document, startup prompts, and guide the new session to pick up exactly where you left off.

## Why

When a coding conversation gets long, LLM context compression degrades quality. You want to start fresh but keep all the knowledge, decisions, and code context. Session Branch solves this by generating a comprehensive handoff document that a new AI session can read to "hot start" your project.

## Features

- **Structured Handoff Document** — 12-section template covering project identity, decision chain, data flow, capability boundary, code changes, platform status, and more
- **Validation Checklist** — 12 categories, 40+ items to ensure not

_meta.json

{
  "ownerId": "kn75zj7vzdyvap84adxa8heyyd82f5eh",
  "slug": "session-branch",
  "version": "1.4.0",
  "publishedAt": 1784083809005
}

references/checklist.md

# Handoff Document Generation Checklist

When generating a handoff document, verify every item is covered.
If an item is not applicable, write "N/A" with a reason.

---

## A. Primary Request and Intent (P2, ~300 tokens)
- [ ] All user intents extracted from user messages
- [ ] Deduplicated and numbered in chronological order

## B. Key Technical Concepts (P2, ~200 tokens)
- [ ] Technical keywords extracted and deduplicated
- [ ] Categorized (Architecture / Tools / Patterns)

## C. Files and Code Sections (P1, ~1200 tokens)
- [ ] All Read/Edit/Write file operations recorded
- [ ] Each file: path + importance + change summary + key code snippet (max 500 chars)
- [ ] Project file structure overview included
- [ ] File paths are project-relative (never absolute with usernames)

## D. Errors and Fixes (P1, ~600 tokens)
- [ ] Recent 5-10 errors recorded (not all)
- [ ] Each error: description + root cause + fix solution

## E. Problem Solving (P2, ~300 tokens)
- [ ] Completed achievements summarized
- [ ] Key decisions documented (chosen vs rejected + reason)

## F. All User Messages (P0, ~800 tokens)
- [ ] All user messages preserved as original text **with mandatory sanitization**
- [ ] Credentials, tokens, API keys, passwords replaced with `<REDACTED_CREDENTIAL>`
- [ ] Real names, emails, phone numbers replaced with `<REDACTED_PII>`
- [ ] Absolute paths with usernames converted to project-relative paths
- [ ] Messages >500 chars truncated with "..."

## G. Conversation Language (P3, ~10 tokens)
- [ ] Primary conversation language detected and recorded

## H. Current Work (P0, ~600 tokens)
- [ ] Active task described (what is being worked on right now)
- [ ] Unfinished Todos listed with status and files involved
- [ ] Branchable directions listed with type, files, prerequisites, complexity

---

## Supplementary: Project Context

### I. Project Identity
- [ ] Project name, description, platforms, version, tech stack, license
- [ ] Project directory is relative path (never absolute)

### J. Platform Status
- [ ] GitHub repo + release status
- [ ] ClawHub version + security status (if applicable)

### K. Environment Variables
- [ ] Env var names listed (yes/no status only)
- [ ] Which need user configuration
- [ ] Token types and required permissions
- [ ] No env var values recorded anywhere

### L. Capability Boundary
- [ ] What the project CAN do
- [ ] What the project CANNOT do
- [ ] Dependency constraints

### M. Knowledge File Index
- [ ] Indexed by scenario (not by filename)
- [ ] Priority annotated (Must-read / On-demand)

### N. User Preferences
- [ ] Preferred language, code style, workflow preferences recorded
- [ ] Only preferences affecting code generation recorded (no personal attributes, no security sensitivity)

---

## O. Startup Prompt
- [ ] Three-step flow: Load -> Report -> Ask
- [ ] Adapted for target IDE platform
- [ ] Role set: "continuation of existing project, not starting from scratch"
- [ ] Project-relative file paths (never

references/handoff-template.md

# Session Handoff Document Template

> This template defines the structure for a project handoff document.
> Based on Trae IDE session-copy architecture: 8 fixed sections with token budget management.
> **Never include personal information** — use placeholders for names, paths, and credentials.

---

## Token Budget Allocation

> Total target: ~4000 tokens. Use the priority table below to compress when content exceeds budget.
>
> **Sanitization applies to ALL sections (including P0)**: Before preserving any user content, strip credentials, tokens, API keys, real names, emails, phone numbers, and personal identifiers. Replace with `<REDACTED>` placeholders. Preservation priority never overrides sanitization — sensitive content is always redacted, even in P0 sections.

| Priority | Section | Token Budget | Compression Rule |
|----------|---------|-------------|-----------------|
| **P0** | All User Messages | ~800 | Preserve original text **with sanitization** (strip credentials/PII, replace with `<REDACTED>`); truncate messages >500 chars with "..." |
| **P0** | Current Work | ~600 | Detailed description of current task state + unfinished Todos **with sanitization** (no token values, no personal paths) |
| **P1** | Files and Code Sections | ~1200 | File paths + modification summary + key code snippets (max 500 chars each) |
| **P1** | Errors and Fixes | ~600 | Error description + solution; keep only recent 5-10 items |
| **P2** | Primary Request and Intent | ~300 | Numbered list, deduplicated |
| **P2** | Key Technical Concepts | ~200 | Keyword list, deduplicated and categorized |
| **P2** | Problem Solving | ~300 | Summary of completed achievements |
| **P3** | Conversation Language | ~10 | Single field |

---

## 1. Primary Request and Intent

> P2 priority | ~300 tokens | Extract from all user messages, deduplicate, number in chronological order

1. `<intent-1>`
2. `<intent-2>`
3. `<intent-3>`

---

## 2. Key Technical Concepts

> P2 priority | ~200 tokens | Extract technical keywords, deduplicate, categorize

- **Architecture**: `<concepts>`
- **Tools/Libraries**: `<concepts>`
- **Patterns**: `<concepts>`

---

## 3. Files and Code Sections

> P1 priority | ~1200 tokens | Record all Read/Edit/Write operations. Keep file path + modification summary + key code snippet (max 500 chars)

| File | Importance | Changes | Key Code Snippet |
|------|-----------|---------|-----------------|
| `<file-1>` | `<why-it-matters>` | `<what-changed>` | `<snippet or "see file">` |
| `<file-2>` | `<why-it-matters>` | `<what-changed>` | `<snippet or "see file">` |

**Project file structure overview**:
```
<project-name>/
├── <file-1>                  # <description>
├── <dir-1>/
│   └── <file>                # <description>
└── <dir-2>/
    └── <file>                # <description>
```

---

## 4. Errors and Fixes

> P1 priority | ~600 tokens | Keep only recent 5-10 errors. Record error description + root cause + solution

| # | Error | Root Cause | Fix |
|---|---

Github ReposUpdated 2d agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.