Skip to content
Back to skills

Save Handoff

ASecurity

Save a concise handoff summary between OpenHands worker agents or when the user explicitly requests a task checkpoint.

  • 182 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 2, 2026
ai-agentsbashnodeapi

Works with

  • cli
  • api
  • mcp

Security analysis

A100/100

Scanned October 2, 2026

npx -y skills add nowledge-co/community --skill save-handoff --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Save Handoff?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Save Handoff
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/nowledge-co-save-handoff-197f5223/badge)](https://www.skillsdirectory.com/skills/nowledge-co-save-handoff-197f5223)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: save-handoff
description: Save a concise handoff summary between OpenHands worker agents or when the user explicitly requests a task checkpoint.
trigger:
  type: keyword
  keywords:
    - handoff
    - checkpoint
    - summarize
    - resume
    - task-complete
---

# Save Handoff

> Persist a compact resumable handoff when an OpenHands agent finishes its phase, or when the user asks for a checkpoint.

## When to Use

**In Multi-Agent Orchestrations:**

- An agent node (e.g. `planner` or `architect`) finishes its scoped task and yields control to the next worker node (`coder` or `reviewer`).
- Capturing intermediate status, open blockers, and concrete next steps.

**On Explicit User Request:**

- "Save a handoff"
- "Checkpoint this work"
- "Leave me a summary of what's done"
- "Remember where we stopped"

**Skip when:**

- Normal intermediate steps within the same agent turn.
- Routine tool executions without meaningful milestone completion.

## Structure of a Handoff

A clean handoff contains:

1. **Objective**: What the agent or canvas node was tasked to do.
2. **Current State**: What was completed, modified files, passing tests.
3. **Decisions Made**: Architectural choices or constraints discovered.
4. **Next Actions**: Exact next steps for the incoming agent node or next session.
5. **Blockers/Flags**: Unresolved items requiring user attention or upstream fix.

## Usage

### Via CLI

```bash
nmem --json t create \
  --title "OpenHands Handoff: API Authentication Refactor" \
  --body "Objective: Refactor JWT middleware\nStatus: Middleware implemented, unit tests passing\nNext: Wire router in server.py" \
  --label openhands --label handoff
```

### Via MCP

Use `memory_add` with `#handoff` label and structured markdown summary:

```json
memory_add({
  "content": "### Handoff: API Authentication Refactor\n\n**Status**: Middleware implemented in `auth.py`.\n**Next**: Wire route guards in `server.py`.\n**Open**: Verify refresh token expiration policy.",
  "labels": ["handoff", "openhands", "checkpoint"]
})
```

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.