一键导入
handover
Generate a handover document for transferring work to another developer or spawning an async agent
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Generate a handover document for transferring work to another developer or spawning an async agent
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Autonomous product factory — /make-a-goal in godmode. Point it at a north-star outcome and it runs the whole delivery machine unattended: brainstorm the feature landscape (Discover), put every idea through a Feasibility Court, cluster into a Roadmap of epics, then per epic Plan → adversarial review → Execute (model-paired build) → VERIFY (drive the real UI/TUI/API, not mocks) → ship a stacked PR, looping until the backlog is dry or a budget/time bound fires. One human gate (roadmap blessing), hard stop rules, always-on tabbed RAG dashboard, crash-proof resume from state files. Use when the user says 'godmode', 'run the factory', 'ship X to nirvana', 'execute this whole backlog autonomously', 'goals within goals', or wants a multi-epic programme driven end-to-end with real end-to-end verification. NOT for single features, one-off verification, or scoped builds — use /make-a-goal, browser/tmux verification skills, or a plain plan for those.
Sync user-level agent config changes back to the ainb-toolkit repo (works for Claude, Codex, Copilot)
Produce a self-contained, richly styled HTML explainer for any topic the user asks about. Picks the right template from a bundled set of 22 visual patterns (feature explainer, concept explainer, module map, PR review, ADR, options paper / trade-off analysis, system diagram, flow- chart, status report, slide deck, prototype, editor, etc.), fills it with real content, augments with inline diagrams via sister skills (/fireworks-tech-graph for architecture / flow / sequence diagrams, /graphify for knowledge graphs), applies a Claude-brand polish layer, and publishes it: by default onto the configured here.now custom domain (path mount + searchable categorised index + password lock per the config's protect rule, driven by ~/.herenow/explainers.json), or a plain here.now URL, or a GitHub gist (--gist / --gist --public). Local-only output is available with --local. Use when Stevie says "/explain-to-me", "explain-to-me X", "make me an explainer for X", "give me an HTML explainer", "render this as a webpage", "ADR for
Self-improvement through conversation analysis. Extracts learnings from corrections and success patterns, permanently encoding them into agent definitions. Philosophy - Correct once, never again.
Claude Code observability skill: analyze session traces stored in Langfuse, extract learnings from corrections, identify success patterns, and propose agent/skill improvements based on historical data. Powers self-improvement through trace analysis of Claude Code sessions.
Create well-formatted git commits for changes made during the session
| name | handover |
| description | Generate a handover document for transferring work to another developer or spawning an async agent |
| user-invocable | true |
Snapshot session state (git, todos, next steps) into a markdown file so another human, a resumed session, or a spawned agent can continue without re-deriving context.
| Want | Use instead |
|---|---|
| Current health / message count only | /health-check |
| Narrative recap of work done | /session-summary |
| Raw session stats | /session-metrics |
| To actually launch the agent now | /spawn-agent (it calls this — Step 5) |
/handover # Standard (human / resume-later)
/handover "notes about current work" # Standard + notes appended
/handover --agent-spawn "task desc" # Agent-spawn (task-focused)
{{TOOL_DIR}} is substituted at deploy time by bootstrap.js to the project-relative tool dir (.claude) — NOT ~/.claude (that is {{HOME_TOOL_DIR}}). Leave both tokens literal — never replace by hand. Then route: standard → Step 2, agent → Step 3.
MODE="standard"; AGENT_TASK=""; NOTES="${1:-}"
if [[ "$1" == "--agent-spawn" ]]; then MODE="agent"; AGENT_TASK="${2:-}"; NOTES=""; fi
DISPLAY_TIME=$(date +"%Y-%m-%d %H:%M:%S")
FILENAME="handover-$(date +"%Y-%m-%d-%H-%M-%S").md"
PRIMARY_LOCATION="{{TOOL_DIR}}/session/${FILENAME}"; BACKUP_LOCATION="./${FILENAME}"
mkdir -p "{{TOOL_DIR}}/session"
Build $CONTENT from this template. Fill every [bracketed] slot from the live session (your todo list, what you were doing) — no brackets may remain. Shell $(...) runs at generation time.
# Handover Document
**Generated**: ${DISPLAY_TIME}
**Session**: $(tmux display-message -p '#S' 2>/dev/null || echo 'unknown')
## Current Work
[What you're working on — 1-3 sentences]
## Task Progress
[Todo list with status: completed / in-progress / pending]
## Technical Context
**Current Branch**: $(git branch --show-current)
**Last Commit**: $(git log -1 --oneline)
**Modified Files**:
$(git status --short)
## Resumption Instructions
1. Review changes: git diff
2. Continue work on [specific task]
3. Test with: [test command]
## Notes
${NOTES}
Cold-handover variant: if a human picks this up days later and needs health status / JIRA ID / phase / progress %, copy assets/template.md and fill its {{...}} placeholders instead. Then go to Step 4.
Task-focused; omit session-health fields (a spawned agent starts fresh). Fill every [bracketed] slot.
# Agent Handover - ${AGENT_TASK}
**Generated**: ${DISPLAY_TIME}
**Parent Session**: $(tmux display-message -p '#S' 2>/dev/null || echo 'unknown')
**Agent Task**: ${AGENT_TASK}
## Context Summary
**Current Work**: [What's in progress]
**Current Branch**: $(git branch --show-current)
**Last Commit**: $(git log -1 --oneline)
## Task Details
**Agent Mission**: ${AGENT_TASK}
**Requirements**:
- [Specific requirements / what needs doing]
**Success Criteria**:
- [How to know when done]
## Technical Context
**Stack**: [Technology stack]
**Key Files**:
$(git status --short)
**Modified Recently**:
$(git log --name-only -5 --oneline)
## Instructions for Agent
1. Review current implementation
2. Make specified changes
3. Add/update tests
4. Verify all tests pass
5. Commit with clear message
## References
**Documentation**: [Links to relevant docs]
**Related Work**: [Related PRs/issues]
The cwd backup survives if {{TOOL_DIR}} is on another volume or gets cleaned.
# Bind the filled template for the active mode into CONTENT first — printf on an
# unset CONTENT writes empty files.
if [ "$MODE" = "agent" ]; then
CONTENT="<the filled Step 3 agent-spawn handover content>"
else
CONTENT="<the filled Step 2 standard handover content>"
fi
printf '%s\n' "$CONTENT" > "$PRIMARY_LOCATION"
printf '%s\n' "$CONTENT" > "$BACKUP_LOCATION"
echo "Handover -> $PRIMARY_LOCATION (backup: $BACKUP_LOCATION)"
Gate — expect both files non-empty.
git branch --show-current empty → detached HEAD / not a repo. Leave Branch blank, continue (do NOT abort).{{TOOL_DIR}}/session exists (Step 1 mkdir) and disk is writable, then retry.Output location (contract): primary {{TOOL_DIR}}/session/handover-{timestamp}.md, backup ./handover-{timestamp}.md.
/spawn-agent ... --with-handover calls this skill in agent mode and copies the result into the agent worktree as .agent-handover.md:
/spawn-agent codex "refactor auth" --with-handover
# Internally calls: /handover --agent-spawn "refactor auth"
Keep the --agent-spawn "<task>" argument shape exactly — spawn-agent depends on it.