[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7E5EBDFYZPMJ1HM484EH8K8":3,"skill-same:01M7E5EBDFYZPMJ1HM484EH8K8":49},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":28,"commitSha":29,"skillDirectory":30,"forksCount":31,"repositoryHtmlUrl":32,"repositoryHomepageUrl":6,"ownerHtmlUrl":33,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":34,"install":35,"resources":48},"01M7E5EBDFYZPMJ1HM484EH8K8","speckit-ext-session-handover",null,"Mid-implementation context handover with proactive monitoring and continuation prompts","stefanwalther\u002Fghn","stefanwalther","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F669728?v=4","01M7E59BWN972ZZWN6SJ3RBZS8","Stefan Walther","01M7E59BX6XCFE72NWR5BN5M78",1,"MIT","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-session-handover\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: speckit-ext-session-handover\ndescription: Mid-implementation context handover with proactive monitoring and continuation prompts\ncompatibility: Requires spec-kit project structure with .specify\u002F directory\nmetadata:\n  author: github-spec-kit\n  source: ext:commands\u002Fspeckit.ext.session-handover.md\n---\n\n## AI Self-Monitoring: Context Awareness\n\n**IMPORTANT**: As the AI assistant, you should proactively monitor context usage during implementation sessions.\n\n### Context Heuristics\n\nMonitor these signals to estimate context usage:\n- **Message count**: More than 30-40 exchanges suggests high context usage\n- **Large file reads**: Each large file (>500 lines) consumes significant context\n- **Cumulative output**: Long responses fill context faster\n- **Tool call count**: Each tool call adds overhead\n\n### Warning Thresholds\n\n| Level | Signals | Action |\n|-------|---------|--------|\n| **Normal** | \u003C30 messages, few large reads | Continue normally |\n| **Warning (~80%)** | 30-40 messages, multiple large reads, noticeable slowdown | Proactively suggest: \"Context is filling up. Consider running `\u002Fspeckit.ext.session-handover` to preserve state.\" |\n| **Critical (~90%)** | >40 messages, many file reads, definite slowdown | Strongly recommend: \"Context nearly full. I recommend we create a handover now to avoid losing implementation context. Run `\u002Fspeckit.ext.session-handover --quick`?\" |\n\n### Proactive Behavior\n\nWhen you sense context pressure:\n1. **Mention it naturally** in your response: \"I notice we've been working for a while and context may be getting full...\"\n2. **Offer the handover option**: \"Would you like me to create a session handover now?\"\n3. **At critical level**: Auto-suggest quick mode to minimize additional context usage\n\n---\n\n## When to Use This Extension\n\n**Use `\u002Fspeckit.ext.session-handover` when:**\n- Context window is running low (AI warns you or you notice slowdown)\n- End of work session (need to pause and continue later)\n- Switching to a different task temporarily\n- Before auto-compact destroys critical implementation context\n\n**Use `\u002Fspeckit.ext.closeout` instead when:**\n- Feature is COMPLETE and ready for PR\n- You want comprehensive COMPLETION.md and HANDOVER.md\n\n**The key difference**: This extension is for PAUSING mid-implementation. Closeout is for FINISHING a complete feature.\n\n---\n\n## User Input\n\n```text\n$ARGUMENTS\n```\n\nYou **MUST** consider the user input before proceeding (if not empty).\n\n## Outline\n\nThis extension captures current implementation state and generates a continuation prompt that enables the next session to pick up immediately without re-reading files.\n\n**Core Principle**: The continuation prompt is the primary deliverable. It must be self-contained, copy-paste ready, and actionable.\n\n### Execution Steps\n\n1. **Parse Arguments and Detect Feature**\n\na. **Parse Arguments**:\n   - Extract spec directory if provided\n   - Check for `--quick` flag (skip all prompts)\n   - Check for `--notes \"...\"` argument\n   - If `--help`: Display usage and exit\n\nb. **Auto-detect Feature** (if no directory provided):\n   - Get current branch: `git rev-parse --abbrev-ref HEAD`\n   - Extract feature number from branch name (pattern: `###-feature-name`)\n   - Find matching spec directory: `specs\u002F{number}-*`\n\nc. **Validate Feature Directory**:\n   - Check spec directory exists\n   - Check tasks.md exists (primary source of progress)\n   - If missing: Use git diff as fallback for state capture\n\nd. **Check if Feature is Complete**:\n   - Read tasks.md and count incomplete tasks\n   - If all tasks complete: Warn \"Feature appears complete. Consider using `\u002Fspeckit.ext.closeout` instead for full completion documentation.\"\n   - Continue if user wants handover anyway\n\n**Display** (unless --quick):\n```\n=== Session Handover ===\n\nFeature: #{number} - {name}\nBranch: {branch-name}\nSpec: specs\u002F{number}-{name}\u002F\n\nThis will:\n1. Capture current implementation state\n2. Create SESSION-{date}.md\n3. Generate continuation prompt for next session\n\nContinue? [Y\u002Fn]\n```\n\n2. **Capture Implementation State**\n\na. **Parse tasks.md**:\n   - Read tasks.md using Read tool\n   - Categorize tasks:\n     - **Completed**: Lines matching `- [x]` or `- [X]`\n     - **In Progress**: First incomplete task after last completed (infer current work)\n     - **Pending**: Remaining incomplete tasks `- [ ]`\n   - Extract task IDs and descriptions\n\nb. **Capture Git State**:\n   - Recent changes: `git diff --name-only HEAD~5..HEAD 2>\u002Fdev\u002Fnull || git diff --name-only`\n   - Modified files: `git status --porcelain`\n   - Current branch: `git rev-parse --abbrev-ref HEAD`\n\nc. **Extract Feature Context**:\n   - Read spec.md for feature title and description\n   - Get current date for timestamp\n\n**Internal State** (don't display, use for generation):\n```\ncompleted_tasks: [T001, T002, ...]\nin_progress_task: T00X\npending_tasks: [T00Y, T00Z, ...]\nrecent_files: [file1.md, file2.py, ...]\nbranch: 012-session-handover\nfeature_name: Session Handover Extension\n```\n\n3. **Gather Developer Notes** (unless --quick)\n\na. **Check for --notes argument**:\n   - If provided: Use the notes value directly\n   - If not provided and not --quick: Prompt for notes\n\nb. **Interactive Notes Prompt** (if not --quick and no --notes):\n   ```\n   Add any notes for the next session? (blockers, decisions, context)\n   \n   Press Enter to skip, or type your notes:\n   > \n   ```\n\nc. **Capture \"What I Was Working On\"**:\n   - AI should describe in its own words what it was actively working on\n   - Be specific: \"Adding the 90% threshold warning\" not \"implementing features\"\n   - Include any pending decisions or open questions\n\n4. **Generate SESSION-{date}.md**\n\nCreate the handover document in the spec directory.\n\n**File**: `specs\u002F{number}-{name}\u002FSESSION-{YYYY-MM-DD}.md`\n\n**If file exists**: Append timestamp to make unique: `SESSION-{YYYY-MM-DD}-{HHMM}.md`\n\n**Content**:\n```markdown\n# Session Handover - {DATE}\n\n**Feature**: #{number} - {name}\n**Branch**: {branch-name}\n**Paused**: {timestamp}\n\n---\n\n## Progress Summary\n\n**Completed**: {count} tasks\n**In Progress**: {count} tasks\n**Remaining**: {count} tasks\n\n### Completed Tasks\n{list of completed tasks with IDs}\n\n### In Progress\n- [ ] {current task ID}: {description} ← CURRENT\n\n### Next Up\n{list of next 3-5 pending tasks}\n\n---\n\n## Recent Changes\n\nFiles modified in this session:\n{list of recently modified files}\n\n---\n\n## What I Was Working On\n\n{AI-generated description of current work - specific and actionable}\n\n---\n\n## Key Decisions Made This Session\n\n{Bullet list of any decisions made, or \"None documented\"}\n\n---\n\n## Developer Notes\n\n{user-provided notes or \"None\"}\n\n---\n\n## Blockers \u002F Questions\n\n{Any unresolved issues, or \"None\"}\n\n---\n\n## Continuation Prompt\n\n{The formatted continuation prompt - see Step 5}\n```\n\n5. **Generate Continuation Prompt**\n\nThis is the PRIMARY DELIVERABLE. Generate a self-contained prompt that enables immediate pickup.\n\n**Format**:\n```\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n📋 CONTINUATION PROMPT - Copy everything below this line\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nContinue implementing {feature name} (#{number}).\n\n## Context\n- **Branch**: `{branch}` (already checked out)\n- **Spec**: `specs\u002F{number}-{name}\u002Fspec.md`\n- **Handover**: `specs\u002F{number}-{name}\u002FSESSION-{date}.md`\n\n## Progress\n- ✅ Completed: {task range or count}\n- 🔄 In Progress: {current task ID} - {description}\n- ⏳ Remaining: {count} tasks\n\n## What I Was Working On\n{specific description of current work}\n\n## Key Decisions Made\n{bullet list of decisions, or \"See handover document\"}\n\n## Next Action\n{specific, actionable instruction for what to do next}\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n```\n\n**Critical Requirements for Continuation Prompt**:\n- **Self-contained**: Next session should NOT need to read spec.md or tasks.md to understand context\n- **Specific Next Action**: Not \"continue implementation\" but \"Add the 90% warning threshold in line 45\"\n- **Copy-paste ready**: Clear delimiters, user copies verbatim\n- **Actionable**: AI can take action in first response, not \"let me read the files first\"\n\n6. **Write and Display**\n\na. **Write SESSION-{date}.md**:\n   - Use Write tool to create the file in spec directory\n   - Include the continuation prompt in the document\n\nb. **Display Summary** (unless --quick):\n   ```\n   === Session Handover Complete ===\n   \n   Feature: #{number} - {name}\n   Branch: {branch}\n   Paused: {timestamp}\n   \n   Progress:\n   ✅ Completed: {count} tasks\n   🔄 In Progress: {current task}\n   ⏳ Remaining: {count} tasks\n   \n   Created:\n   📄 specs\u002F{number}-{name}\u002FSESSION-{date}.md\n   \n   ---\n   ```\n\nc. **Display Continuation Prompt**:\n   - Show the formatted continuation prompt prominently\n   - This is what the user copies for the next session\n\nd. **Final Note**:\n   ```\n   Copy the continuation prompt above to start your next session.\n   The next AI will have full context to continue immediately.\n   ```\n\n## Quick Mode Behavior\n\nWhen `--quick` flag is used OR at critical context threshold:\n\n1. **Skip all prompts**: No confirmation, no interactive notes\n2. **Minimal output**: Write file, display only the continuation prompt\n3. **Fast execution**: Minimize token usage\n4. **Still capture state**: Full state capture, just no interaction\n\n**Quick mode output**:\n```\nSession handover created: specs\u002F{number}-{name}\u002FSESSION-{date}.md\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n📋 CONTINUATION PROMPT\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n{continuation prompt}\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n```\n\n## Usage Examples\n\n```bash\n# Standard handover (auto-detect feature)\n\u002Fspeckit.ext.session-handover\n\n# Quick mode - no prompts, minimal output\n\u002Fspeckit.ext.session-handover --quick\n\n# With notes\n\u002Fspeckit.ext.session-handover --notes \"Was debugging the threshold logic\"\n\n# Specific feature directory\n\u002Fspeckit.ext.session-handover specs\u002F012-session-handover\u002F\n\n# Combine flags\n\u002Fspeckit.ext.session-handover --quick --notes \"Context critical\"\n```\n\n## Edge Cases\n\n- **No tasks.md**: Use git diff as primary state source, warn user\n- **No git changes**: Still create handover with tasks state\n- **All tasks complete**: Warn and suggest closeout instead\n- **Multiple sessions same day**: Append timestamp to filename\n- **Auto-compact already happened**: Capture what's available, warn about potential context loss\n\n## Notes\n\n- **Primary Deliverable**: The continuation prompt, not the SESSION file\n- **Goal**: Next session acts immediately, no \"let me read the files first\"\n- **Context Efficiency**: Quick mode uses minimal tokens\n- **Preservation**: SESSION-{date}.md files should be kept (valuable for project history)\n- **Platform**: Works on Claude Code and Cursor (AI self-monitoring)","## AI Self-Monitoring: Context Awareness\n\n**IMPORTANT**: As the AI assistant, you should proactively monitor context usage during implementation sessions.\n\n### Context Heuristics\n\nMonitor these signals to estimate context usage:\n- **Message count**: More than 30-40 exchanges suggests high context usage\n- **Large file reads**: Each large file (>500 lines) consumes significant context\n- **Cumulative output**: Long responses fill context faster\n- **Tool call count**: Each tool call adds overhead\n\n### Warning Thresholds\n\n| Level | Signals | Action |\n|-------|---------|--------|\n| **Normal** | \u003C30 messages, few large reads | Continue normally |\n| **Warning (~80%)** | 30-40 messages, multiple large reads, noticeable slowdown | Proactively suggest: \"Context is filling up. Consider running `\u002Fspeckit.ext.session-handover` to preserve state.\" |\n| **Critical (~90%)** | >40 messages, many file reads, definite slowdown | Strongly recommend: \"Context nearly full. I recommend we create a handover now to avoid losing implementation context. Run `\u002Fspeckit.ext.session-handover --quick`?\" |\n\n### Proactive Behavior\n\nWhen you sense context pressure:\n1. **Mention it naturally** in your response: \"I notice we've been working for a while and context may be getting full...\"\n2. **Offer the handover option**: \"Would you like me to create a session handover now?\"\n3. **At critical level**: Auto-suggest quick mode to minimize additional context usage\n\n---\n\n## When to Use This Extension\n\n**Use `\u002Fspeckit.ext.session-handover` when:**\n- Context window is running low (AI warns you or you notice slowdown)\n- End of work session (need to pause and continue later)\n- Switching to a different task temporarily\n- Before auto-compact destroys critical implementation context\n\n**Use `\u002Fspeckit.ext.closeout` instead when:**\n- Feature is COMPLETE and ready for PR\n- You want comprehensive COMPLETION.md and HANDOVER.md\n\n**The key difference**: This extension is for PAUSING mid-implementation. Closeout is for FINISHING a complete feature.\n\n---\n\n## User Input\n\n```text\n$ARGUMENTS\n```\n\nYou **MUST** consider the user input before proceeding (if not empty).\n\n## Outline\n\nThis extension captures current implementation state and generates a continuation prompt that enables the next session to pick up immediately without re-reading files.\n\n**Core Principle**: The continuation prompt is the primary deliverable. It must be self-contained, copy-paste ready, and actionable.\n\n### Execution Steps\n\n1. **Parse Arguments and Detect Feature**\n\na. **Parse Arguments**:\n   - Extract spec directory if provided\n   - Check for `--quick` flag (skip all prompts)\n   - Check for `--notes \"...\"` argument\n   - If `--help`: Display usage and exit\n\nb. **Auto-detect Feature** (if no directory provided):\n   - Get current branch: `git rev-parse --abbrev-ref HEAD`\n   - Extract feature number from branch name (pattern: `###-feature-name`)\n   - Find matching spec directory: `specs\u002F{number}-*`\n\nc. **Validate Feature Directory**:\n   - Check spec directory exists\n   - Check tasks.md exists (primary source of progress)\n   - If missing: Use git diff as fallback for state capture\n\nd. **Check if Feature is Complete**:\n   - Read tasks.md and count incomplete tasks\n   - If all tasks complete: Warn \"Feature appears complete. Consider using `\u002Fspeckit.ext.closeout` instead for full completion documentation.\"\n   - Continue if user wants handover anyway\n\n**Display** (unless --quick):\n```\n=== Session Handover ===\n\nFeature: #{number} - {name}\nBranch: {branch-name}\nSpec: specs\u002F{number}-{name}\u002F\n\nThis will:\n1. Capture current implementation state\n2. Create SESSION-{date}.md\n3. Generate continuation prompt for next session\n\nContinue? [Y\u002Fn]\n```\n\n2. **Capture Implementation State**\n\na. **Parse tasks.md**:\n   - Read tasks.md using Read tool\n   - Categorize tasks:\n     - **Completed**: Lines matching `- [x]` or `- [X]`\n     - **In Progress**: First incomplete task after last completed (infer current work)\n     - **Pending**: Remaining incomplete tasks `- [ ]`\n   - Extract task IDs and descriptions\n\nb. **Capture Git State**:\n   - Recent changes: `git diff --name-only HEAD~5..HEAD 2>\u002Fdev\u002Fnull || git diff --name-only`\n   - Modified files: `git status --porcelain`\n   - Current branch: `git rev-parse --abbrev-ref HEAD`\n\nc. **Extract Feature Context**:\n   - Read spec.md for feature title and description\n   - Get current date for timestamp\n\n**Internal State** (don't display, use for generation):\n```\ncompleted_tasks: [T001, T002, ...]\nin_progress_task: T00X\npending_tasks: [T00Y, T00Z, ...]\nrecent_files: [file1.md, file2.py, ...]\nbranch: 012-session-handover\nfeature_name: Session Handover Extension\n```\n\n3. **Gather Developer Notes** (unless --quick)\n\na. **Check for --notes argument**:\n   - If provided: Use the notes value directly\n   - If not provided and not --quick: Prompt for notes\n\nb. **Interactive Notes Prompt** (if not --quick and no --notes):\n   ```\n   Add any notes for the next session? (blockers, decisions, context)\n   \n   Press Enter to skip, or type your notes:\n   > \n   ```\n\nc. **Capture \"What I Was Working On\"**:\n   - AI should describe in its own words what it was actively working on\n   - Be specific: \"Adding the 90% threshold warning\" not \"implementing features\"\n   - Include any pending decisions or open questions\n\n4. **Generate SESSION-{date}.md**\n\nCreate the handover document in the spec directory.\n\n**File**: `specs\u002F{number}-{name}\u002FSESSION-{YYYY-MM-DD}.md`\n\n**If file exists**: Append timestamp to make unique: `SESSION-{YYYY-MM-DD}-{HHMM}.md`\n\n**Content**:\n```markdown\n# Session Handover - {DATE}\n\n**Feature**: #{number} - {name}\n**Branch**: {branch-name}\n**Paused**: {timestamp}\n\n---\n\n## Progress Summary\n\n**Completed**: {count} tasks\n**In Progress**: {count} tasks\n**Remaining**: {count} tasks\n\n### Completed Tasks\n{list of completed tasks with IDs}\n\n### In Progress\n- [ ] {current task ID}: {description} ← CURRENT\n\n### Next Up\n{list of next 3-5 pending tasks}\n\n---\n\n## Recent Changes\n\nFiles modified in this session:\n{list of recently modified files}\n\n---\n\n## What I Was Working On\n\n{AI-generated description of current work - specific and actionable}\n\n---\n\n## Key Decisions Made This Session\n\n{Bullet list of any decisions made, or \"None documented\"}\n\n---\n\n## Developer Notes\n\n{user-provided notes or \"None\"}\n\n---\n\n## Blockers \u002F Questions\n\n{Any unresolved issues, or \"None\"}\n\n---\n\n## Continuation Prompt\n\n{The formatted continuation prompt - see Step 5}\n```\n\n5. **Generate Continuation Prompt**\n\nThis is the PRIMARY DELIVERABLE. Generate a self-contained prompt that enables immediate pickup.\n\n**Format**:\n```\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n📋 CONTINUATION PROMPT - Copy everything below this line\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nContinue implementing {feature name} (#{number}).\n\n## Context\n- **Branch**: `{branch}` (already checked out)\n- **Spec**: `specs\u002F{number}-{name}\u002Fspec.md`\n- **Handover**: `specs\u002F{number}-{name}\u002FSESSION-{date}.md`\n\n## Progress\n- ✅ Completed: {task range or count}\n- 🔄 In Progress: {current task ID} - {description}\n- ⏳ Remaining: {count} tasks\n\n## What I Was Working On\n{specific description of current work}\n\n## Key Decisions Made\n{bullet list of decisions, or \"See handover document\"}\n\n## Next Action\n{specific, actionable instruction for what to do next}\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n```\n\n**Critical Requirements for Continuation Prompt**:\n- **Self-contained**: Next session should NOT need to read spec.md or tasks.md to understand context\n- **Specific Next Action**: Not \"continue implementation\" but \"Add the 90% warning threshold in line 45\"\n- **Copy-paste ready**: Clear delimiters, user copies verbatim\n- **Actionable**: AI can take action in first response, not \"let me read the files first\"\n\n6. **Write and Display**\n\na. **Write SESSION-{date}.md**:\n   - Use Write tool to create the file in spec directory\n   - Include the continuation prompt in the document\n\nb. **Display Summary** (unless --quick):\n   ```\n   === Session Handover Complete ===\n   \n   Feature: #{number} - {name}\n   Branch: {branch}\n   Paused: {timestamp}\n   \n   Progress:\n   ✅ Completed: {count} tasks\n   🔄 In Progress: {current task}\n   ⏳ Remaining: {count} tasks\n   \n   Created:\n   📄 specs\u002F{number}-{name}\u002FSESSION-{date}.md\n   \n   ---\n   ```\n\nc. **Display Continuation Prompt**:\n   - Show the formatted continuation prompt prominently\n   - This is what the user copies for the next session\n\nd. **Final Note**:\n   ```\n   Copy the continuation prompt above to start your next session.\n   The next AI will have full context to continue immediately.\n   ```\n\n## Quick Mode Behavior\n\nWhen `--quick` flag is used OR at critical context threshold:\n\n1. **Skip all prompts**: No confirmation, no interactive notes\n2. **Minimal output**: Write file, display only the continuation prompt\n3. **Fast execution**: Minimize token usage\n4. **Still capture state**: Full state capture, just no interaction\n\n**Quick mode output**:\n```\nSession handover created: specs\u002F{number}-{name}\u002FSESSION-{date}.md\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n📋 CONTINUATION PROMPT\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n{continuation prompt}\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n```\n\n## Usage Examples\n\n```bash\n# Standard handover (auto-detect feature)\n\u002Fspeckit.ext.session-handover\n\n# Quick mode - no prompts, minimal output\n\u002Fspeckit.ext.session-handover --quick\n\n# With notes\n\u002Fspeckit.ext.session-handover --notes \"Was debugging the threshold logic\"\n\n# Specific feature directory\n\u002Fspeckit.ext.session-handover specs\u002F012-session-handover\u002F\n\n# Combine flags\n\u002Fspeckit.ext.session-handover --quick --notes \"Context critical\"\n```\n\n## Edge Cases\n\n- **No tasks.md**: Use git diff as primary state source, warn user\n- **No git changes**: Still create handover with tasks state\n- **All tasks complete**: Warn and suggest closeout instead\n- **Multiple sessions same day**: Append timestamp to filename\n- **Auto-compact already happened**: Capture what's available, warn about potential context loss\n\n## Notes\n\n- **Primary Deliverable**: The continuation prompt, not the SESSION file\n- **Goal**: Next session acts immediately, no \"let me read the files first\"\n- **Context Efficiency**: Quick mode uses minimal tokens\n- **Preservation**: SESSION-{date}.md files should be kept (valuable for project history)\n- **Platform**: Works on Claude Code and Cursor (AI self-monitoring)",{"name":5,"metadata":24,"description":7,"compatibility":27},{"author":25,"source":26},"github-spec-kit","ext:commands\u002Fspeckit.ext.session-handover.md","Requires spec-kit project structure with .specify\u002F directory","main","2045797696f3923d9e47519e2302e944260b2266",".specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-session-handover",0,"https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn","https:\u002F\u002Fgithub.com\u002Fstefanwalther","2026-11-15 06:19:35",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add stefanwalther\u002Fghn --skill speckit-ext-session-handover",{"tool":41,"command":42},"bunx","bunx skills add stefanwalther\u002Fghn --skill speckit-ext-session-handover",{"tool":44,"command":45},"pnpm","pnpm dlx skills add stefanwalther\u002Fghn --skill speckit-ext-session-handover","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-session-handover\u002FSKILL.md",true,[],[50,57,64,71,78,85],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":19},"01M7E5EJAJD6R57EE4YSMDJWMP","speckit-ext-validate","Pre-merge validation and quality gate for feature completion","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-validate\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":19},"01M7E5EFW37C0QJ0VZZG7SAWHD","speckit-ext-update-workspace","Update workspace documentation after feature completion","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-update-workspace\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7E5E92Y08ME10QRAW8T0ES9","speckit-ext-scope-change","Document scope changes and ad-hoc additions during feature implementation","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-scope-change\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7E5E3DJ0B3B0FHF0YQAE2JV","speckit-ext-pr","Create a pull request for spec-kit features with agent-based workflow and spec-kit context awareness","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-pr\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7E5E1B26HH69228R8MD54H5","speckit-ext-post-merge","Spec-kit workflow command: speckit-ext-post-merge","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-post-merge\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7E5DXC0ZRPQW9J7T163FJNM","speckit-ext-manual-testing","User-led manual testing with structured feedback collection","https:\u002F\u002Fgithub.com\u002Fstefanwalther\u002Fghn\u002Fblob\u002Fmain\u002F.specify\u002Fextensions\u002Fext\u002F.specify-dev\u002Fagent-commands\u002Fclaude\u002Fspeckit-ext-manual-testing\u002FSKILL.md",[],[]]