# Agentpack Handoff

> Package current work for another real agent session.

- Skill: `vishal2612200/agentpack-handoff` (Agent Skill)
- Install (CLI): `npx skillmds@latest add vishal2612200/agentpack-handoff`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vishal2612200/agentpack-handoff/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- License: AGPL-3.0-only
- Author: vishal2612200 (https://skillmd.com/u/vishal2612200)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/vishal2612200/agentpack-handoff

---


# AgentPack Handoff

Use when the user invokes `$agentpack-handoff [name]` in Codex or `/agentpack-handoff [name]` in Claude Code.

Inspect the current task, acceptance criteria, work completed and remaining, decisions, blockers, next action, Git-visible changes, and real validation evidence. Create the mandatory structured report, then call MCP `create_handoff(report=..., name="<requested name>")`. If MCP is unavailable, run `agentpack handoff create --input <report.json> --name "<requested name>"`.

Use `not_run` plus a reason for validation that did not run. Do not claim checks passed without evidence. Return the generated memorable name, never the internal handoff UUID.