/handoff-continue
Create a handoff document and automatically spawn a new Claude session in a Zellij pane to continue the work seamlessly.
$ npx -y skills add qdhenry/Claude-Command-Suite --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/handoff-continue
Context preview
What this command does when you run it.
Create a handoff document and automatically spawn a new Claude session in a Zellij pane to continue the work seamlessly.
Command definition
handoff-continue.mdname: handoff-continue
model: opus
description: Create a handoff document and automatically spawn a new Claude session in a Zellij pane to continue the work seamlessly.
author: Quintin Henry (https://github.com/qdhenry/)
Context Window Handoff with Auto-Continue
Create a comprehensive handoff document and automatically spawn a new Claude Code session in a Zellij pane to continue the work without manual intervention.
Arguments
Optional argument: filename for the handoff document (without path)
- Example: `/handoff-continue feature-auth-handoff.md`
- If not provided, auto-generates: `HANDOFF_[TOPIC]_[TIMESTAMP].md`
Role
You are a senior engineer creating a comprehensive handoff document AND orchestrating a seamless transition to a new Claude session. You must capture ALL relevant context, then spawn a new session that can immediately continue the work.
Task
1. Analyze the entire conversation to extract key work context 2. Generate a comprehensive handoff document 3. Save it to the specified location (or auto-generate filename) 4. **Spawn a new Zellij pane with Claude Code loaded with the handoff document**
Instructions
Phase 1: Context Gathering
Analyze the conversation for:
- What task/feature/bug was being worked on
- All files that were read, created, or modified
- Key decisions made and their rationale
- Problems encountered and how they were solved
- Current state of the work (what's done vs pending)
- Any open questions or blockers
Phase 2: Git Analysis
If in a git repository, capture:
- Current branch name
- Uncommitted changes (staged and unstaged)
- Recent commits made during this session
- Files in .gitignore that might be relevant
Phase 3: Document Generation
Create a structured document with all sections below.
Phase 4: Spawn Continuation Session
After saving the handoff document, execute the Zellij command to spawn a new pane:
# Create a new Zellij pane and launch Claude with the handoff document
# IMPORTANT: Use full path to claude executable (not alias) and omit -c flag to keep pane open
zellij run -n "Claude Continue" -- bash -c "cd '[PROJECT_DIR]' && /Users/quintinhenry/.claude/local/claude --dangerously-skip-permissions '[HANDOFF_FILE_PATH]'"
**IMPORTANT:**
- Replace `[PROJECT_DIR]` with the actual project directory path
- Replace `[HANDOFF_FILE_PATH]` with the actual absolute path to the saved handoff document
- Must use full path `/Users/quintinhenry/.claude/local/claude` (not `claude` alias) because new shell doesn't inherit aliases
- Omit the `-c` flag so the pane stays open if Claude exits
Output Format
# Handoff: [Brief Title of Work]
**Created:** [timestamp]
**Branch:** [current branch]
**Session Duration:** [approximate time if determinable]
**Continuation:** A new Claude session will be spawned automatically
---
## Summary
[2-3 sentence executive summary of what was being worked on and current state]
---
## Work Completed
### Changes Made
- [ ] [Specific change 1 - use checkbox format for easy verification]
- [ ] [Specific change 2]
- [ ] [Specific change 3]
### Key Decisions
| Decision | Rationale | Alternatives Considered |
| ------------ | --------------------- | -------------------------- |
| [Decision 1] | [Why this was chosen] | [What else was considered] |
| [Decision 2] | [Why this was chosen] | [What else was considered] |
---
## Files Affected
### Created
- `path/to/new/file.ext` - [purpose/description]
### Modified
- `path/to/modified/file.ext` - [what was changed and why]
- Lines/functions affected: [specifics]
### Read (Reference)
- `path/to/reference/file.ext` - [why it was referenced]
### Deleted
- `path/to/deleted/file.ext` - [why it was removed]
---
## Technical Context
### Architecture/Design Notes
[Any architectural decisions, patterns used, or design considerations]
### Dependencies
- [New dependencies added, if any]
- [External services or APIs used]
### Configuration Changes
- [Environment variables, config files, etc.]
---
## Things to Know
### Gotchas & Pitfalls
- [Non-obvious behavior or edge cases discovered]
- [Potential issues to watch out for]
### Assumptions Made
- [Any assumptions that were made during development]
### Known Issues
- [Any issues that were discovered but not fixed]
- [Technical debt introduced]
---
## Current State
### What's Working
- [Feature/component 1 - status]
- [Feature/component 2 - status]
### What's Not Working
- [Issue 1 - description and suspected cause]
- [Issue 2 - description and suspected cause]
### Tests
- [ ] Unit tests: [passing/failing/not written]
- [ ] Integration tests: [status]
- [ ] Manual testing: [what was tested]
---
## Next Steps
### Immediate (Start Here)
1. [Most critical next action with specific details]
2. [Second priority action]
3. [Third priority action]
### Subsequent
- [Lower priority items]
- [Nice-to-haves]
### Blocked On
- [Any blockers with context on how to unblock]
---
## Related Resources
### Documentation
- [Links to relevant docs, PRs, issues]
### Commands to Run
```bash
# Useful commands for continuing this work
[command 1]
[command 2]
### Search Queries
If you need to find more context:
- `[grep/search pattern 1]` - finds [what]
- `[grep/search pattern 2]` - finds [what]
---
## Open Questions
- [ ] [Question 1 that needs answering]
- [ ] [Question 2 that needs answering]
---
## Session Notes
[Any additional context, observations, or notes that don't fit above]
---
_This handoff was generated at context window capacity. A new Claude session is being spawned to continue._
````
## Constraints
- Be comprehensive but concise - prioritize actionable information
- Use specific file paths, function names, and line numbers where relevant
- Write in present tense for current state, future tense for next steps
- Include code snippets only when absolutely necessary for context
- Maximum 200
Read more
name: handoff-continue model: opus description: Create a handoff document and automatically spawn a new Claude session in a Zellij pane to continue the work seamlessly. author: Quintin Henry (https://github.com/qdhenry/)
Context Window Handoff with Auto-Continue
Create a comprehensive handoff document and automatically spawn a new Claude Code session in a Zellij pane to continue the work without manual intervention.
Arguments
Optional argument: filename for the handoff document (without path)
- Example: `/handoff-continue feature-auth-handoff.md`
- If not provided, auto-generates: `HANDOFF_[TOPIC]_[TIMESTAMP].md`
Role
You are a senior engineer creating a comprehensive handoff document AND orchestrating a seamless transition to a new Claude session. You must capture ALL relevant context, then spawn a new session that can immediately continue the work.
Task
1. Analyze the entire conversation to extract key work context 2. Generate a comprehensive handoff document 3. Save it to the specified location (or auto-generate filename) 4. **Spawn a new Zellij pane with Claude Code loaded with the handoff document**
Instructions
Phase 1: Context Gathering
Analyze the conversation for:
- What task/feature/bug was being worked on
- All files that were read, created, or modified
- Key decisions made and their rationale
- Problems encountered and how they were solved
- Current state of the work (what's done vs pending)
- Any open questions or blockers
Phase 2: Git Analysis
If in a git repository, capture:
- Current branch name
- Uncommitted changes (staged and unstaged)
- Recent commits made during this session
- Files in .gitignore that might be relevant
Phase 3: Document Generation
Create a structured document with all sections below.
Phase 4: Spawn Continuation Session
After saving the handoff document, execute the Zellij command to spawn a new pane:
# Create a new Zellij pane and launch Claude with the handoff document # IMPORTANT: Use full path to claude executable (not alias) and omit -c flag to keep pane open zellij run -n "Claude Continue" -- bash -c "cd '[PROJECT_DIR]' && /Users/quintinhenry/.claude/local/claude --dangerously-skip-permissions '[HANDOFF_FILE_PATH]'"
**IMPORTANT:**
- Replace `[PROJECT_DIR]` with the actual project directory path
- Replace `[HANDOFF_FILE_PATH]` with the actual absolute path to the saved handoff document
- Must use full path `/Users/quintinhenry/.claude/local/claude` (not `claude` alias) because new shell doesn't inherit aliases
- Omit the `-c` flag so the pane stays open if Claude exits
Output Format
# Handoff: [Brief Title of Work] **Created:** [timestamp] **Branch:** [current branch] **Session Duration:** [approximate time if determinable] **Continuation:** A new Claude session will be spawned automatically --- ## Summary [2-3 sentence executive summary of what was being worked on and current state] --- ## Work Completed ### Changes Made - [ ] [Specific change 1 - use checkbox format for easy verification] - [ ] [Specific change 2] - [ ] [Specific change 3] ### Key Decisions | Decision | Rationale | Alternatives Considered | | ------------ | --------------------- | -------------------------- | | [Decision 1] | [Why this was chosen] | [What else was considered] | | [Decision 2] | [Why this was chosen] | [What else was considered] | --- ## Files Affected ### Created - `path/to/new/file.ext` - [purpose/description] ### Modified - `path/to/modified/file.ext` - [what was changed and why] - Lines/functions affected: [specifics] ### Read (Reference) - `path/to/reference/file.ext` - [why it was referenced] ### Deleted - `path/to/deleted/file.ext` - [why it was removed] --- ## Technical Context ### Architecture/Design Notes [Any architectural decisions, patterns used, or design considerations] ### Dependencies - [New dependencies added, if any] - [External services or APIs used] ### Configuration Changes - [Environment variables, config files, etc.] --- ## Things to Know ### Gotchas & Pitfalls - [Non-obvious behavior or edge cases discovered] - [Potential issues to watch out for] ### Assumptions Made - [Any assumptions that were made during development] ### Known Issues - [Any issues that were discovered but not fixed] - [Technical debt introduced] --- ## Current State ### What's Working - [Feature/component 1 - status] - [Feature/component 2 - status] ### What's Not Working - [Issue 1 - description and suspected cause] - [Issue 2 - description and suspected cause] ### Tests - [ ] Unit tests: [passing/failing/not written] - [ ] Integration tests: [status] - [ ] Manual testing: [what was tested] --- ## Next Steps ### Immediate (Start Here) 1. [Most critical next action with specific details] 2. [Second priority action] 3. [Third priority action] ### Subsequent - [Lower priority items] - [Nice-to-haves] ### Blocked On - [Any blockers with context on how to unblock] --- ## Related Resources ### Documentation - [Links to relevant docs, PRs, issues] ### Commands to Run ```bash # Useful commands for continuing this work [command 1] [command 2]
### Search Queries If you need to find more context: - `[grep/search pattern 1]` - finds [what] - `[grep/search pattern 2]` - finds [what] --- ## Open Questions - [ ] [Question 1 that needs answering] - [ ] [Question 2 that needs answering] --- ## Session Notes [Any additional context, observations, or notes that don't fit above] --- _This handoff was generated at context window capacity. A new Claude session is being spawned to continue._ ```` ## Constraints - Be comprehensive but concise - prioritize actionable information - Use specific file paths, function names, and line numbers where relevant - Write in present tense for current state, future tense for next steps - Include code snippets only when absolutely necessary for context - Maximum 200
A comprehensive development toolkit designed following Anthropic's Claude Code Best Practices for AI-assisted software development.
Repo: qdhenry/Claude-Command-Suite
Other commands on claude-command-suite.
- /boundary-bbcr-fallback
Execute automatic BBCR (Collapse-Rebirth Correction) when knowledge boundaries are exceeded or reasoning fails.
Open command - /boundary-detect
Analyze semantic position relative to knowledge boundaries to prevent hallucination and identify uncertainty zones.
Open command - /boundary-heatmap
Generate a visual heatmap of knowledge boundaries showing safe zones, risk areas, and semantic coverage.
Open command - /boundary-risk-assess
Evaluate the current risk level and provide detailed analysis of potential hallucination or reasoning failure.
Open command - /boundary-safe-bridge
Find and construct semantic bridges to safely navigate from current position to target concept without crossing dangerous boundaries.
Open command - /optimize-prompt
Takes an input prompt and returns ONLY a token-optimized version that preserves meaning while minimizing token count. Based on LLM tokenization principles: common words tokenize more efficiently, unusual words break into more tokens, and conciseness reduces cost.
Open command

