todero-hand-off-a-task

Structures task hand-offs between agents using five required context fields.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/nabitllc/todero --skill todero-hand-off-a-task-nabitllc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: todero-hand-off-a-task
Source: https://github.com/nabitllc/todero/tree/main/skills/todero-hand-off-a-task
Command: npx skills add https://github.com/nabitllc/todero --skill todero-hand-off-a-task-nabitllc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When one agent hands a task to another, critical context gets lost: the goal, the definition of done, what the requester actually said, and prior reviewer verdicts. This Skill defines a five-field hand-off format so the next agent picks up work with full context instead of guessing. ## Core Features & Use Cases - Five-field hand-off checklist: Goal, Feature, Done-when criteria, the requester's actual words, and the last reviewer verdict. - Relay rule for claims: Anything relayed from an earlier agent is marked as a CLAIM to verify, preventing inferences from being passed off as decisions. - Scope definition: Names the expected output shape, sources to use, and task boundaries so two agents do not duplicate work. - Use Case: In a multi-agent Todero workflow, a planning agent finishes its portion of a feature and hands the task to a drafting agent, which reads the five fields and the prior reviewer's verdict before starting. ## Quick Start Ask the agent to hand off its current task using the five-field format including the goal, done-when criteria, and any prior reviewer verdict.

Frequently Asked Questions about todero-hand-off-a-task

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

FAQPage Schema
How do I hand off a task between AI agents without losing context?

Include five fields in the hand-off: the goal the task serves, the feature it belongs to, a one-line done-when criterion, the requester's actual words, and the last reviewer verdict if one exists. This gives the receiving agent full context before starting.

What information should a task hand-off include for multi-agent workflows?

A hand-off should state the goal, the plan feature, the completion criterion, a quote or paraphrase of what the requester said, and any prior reviewer verdict. It should also name the expected output shape, sources to use, and task boundaries.

How do I prevent agents from duplicating work in a multi-agent system?

Define the task's scope explicitly: name the expected output shape, the sources to use, and the edge of the work. Vague task labels are the main cause of two agents doing the same work twice.

Why mark relayed information as a claim in agent hand-offs?

Marking relayed statements as claims prevents an inference about what the owner wants from being treated as a stated decision. The receiving agent is told to verify the claim before acting rather than trusting it as fact.

When should the last reviewer verdict be added to a task?

Add the verdict whenever a reviewer has already evaluated the work before the hand-off. State it in one sentence so the next agent checks the feedback before starting instead of repeating rejected approaches.