skills.rest
allemaar
allemaarCOMMUNITY
Explore all skills by this author →
Website𝕏@allemaarGitHub

handoff

Generates structured handoff briefs with YAML frontmatter and YON blocks for cold agent sessions.

14|1|Updated Jun 14, 2026
Search Tags:#agent orchestration#obsidian#handoff#session continuity#task delegation#yaml frontmatter#yon
One-click install
npx skills add https://github.com/allemaar/open-skills --skill handoff-allemaar
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: handoff
Source: https://github.com/allemaar/open-skills/tree/main/skills/handoff
Command: npx skills add https://github.com/allemaar/open-skills --skill handoff-allemaar

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Transferring work to a fresh AI agent session usually loses critical context, acceptance criteria, and verification steps, leaving the receiving agent to guess. This Skill packages a task into a complete, self-contained handoff brief that any new LLM session can pick up and execute autonomously. ## Core Features & Use Cases - Structured brief generation: Produces an Obsidian-compatible markdown file with YAML frontmatter and YON-tagged blocks (@TASK, @OUTCOME, @SOURCES, @SCOPE) saved to a vault path. - Three dispatch profiles: Supports solo agent dispatch, orchestrator dispatch with validated subtask decomposition, and new-session continuation that auto-collects in-flight git state. - Verification discipline: Requires verifiable shell commands and mandatory source lists, pushing back on under-specified tasks, and emits a copy-paste activation phrase for the receiving session. - Use Case: You have a bug fix that needs a fresh agent with clean context. Run /handoff, answer the prompts for task, scope, evidence, and acceptance commands, and receive a vault file plus a one-line activation phrase to paste into the new session. ## Quick Start Ask the agent to create a handoff for your task by running /handoff and answering the prompts for task description, scope paths, evidence, and acceptance verification commands.

Frequently Asked Questions about handoff

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I hand off a task to a fresh AI agent session?▼

Run /handoff and provide the task description, in-scope absolute paths, evidence pointer, acceptance sentence, and verifiable shell commands. The skill writes a structured brief file to your vault and emits a one-line activation phrase to paste into the new session.

What is the difference between handoff and orchestrate-mode?▼

The handoff skill produces a portable brief artifact consumed by a separate cold session, while orchestrate-mode keeps the current session as coordinator with no artifact produced. Use handoff for dispatching work away; use orchestrate-mode to manage workers in place.

Can a handoff brief work with agents other than Claude Code?▼

Yes, briefs are runtime-agnostic by design. The skill forbids prescribing runtime-specific tool names in the brief body and describes outcomes instead, so Codex or other LLM runtimes can execute the same brief.

What happens if a handoff has no required sources listed?▼

The skill pushes back and asks you to either name required sources or explicitly confirm none are needed. Confirmed empty sources drop the receiving executor into advisory mode with a prominent warning and no source-ingestion gate.

Why does the handoff skill reject tasks without verify commands?▼

Handoffs without measurable acceptance are unmeasurable, so the skill requires at least one verifiable shell command in the verify field. If you cannot name one, it pushes back to reconsider the task scope before writing the brief.

Related Skills

Explore More Software Engineering Skills→
TeamSloppyTeamSloppy
26

mode-plan

Generate executable software-engineering plans without altering source code.

Official
Advanced
dxiirendxiiren

claude-transfer

Writes pointer-based session handoff briefs for resuming work in fresh Claude sessions.

Community
Intermediate
imaginationeverywhereimaginationeverywhere

cursor-orchestration

Coordinates parallel Cursor Agent CLI sessions with configurable prompts, modes, and models for auditable output.

Official
Advanced
H2OSLabsH2OSLabs

elfiee-fe-dev

Guide React, TypeScript, Vite, and Zustand frontend development with Tauri IPC.

Official
Intermediate