skills.rest
rob-broadley
rob-broadleyCOMMUNITY
Explore all skills by this author →
GitHub

sub-agent-patterns

Define agent-to-agent handoff patterns with structured completion reports.

Updated May 10, 2026
Search Tags:#AI agents#orchestration#agent handoff#autonomous execution#completion report#clarification protocol
One-click install
npx skills add https://github.com/rob-broadley/ai-airbase --skill sub-agent-patterns-rob-broadley
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sub-agent-patterns
Source: https://github.com/rob-broadley/ai-airbase/tree/main/cadre/skills/sub-agent-patterns
Command: npx skills add https://github.com/rob-broadley/ai-airbase --skill sub-agent-patterns-rob-broadley

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill standardizes how an orchestrator-to-sub-agent handoff should detect handoff mode, execute autonomously, ask for clarification internally, and report results in a way the caller can continue orchestration safely.

Core Features & Use Cases

  • Handoff mode detection: Interprets agent-tool invocations structured as Task / Context / Constraints / Success criteria, especially when the prompt references an orchestrating agent like mission-control.
  • Autonomous execution in handoff mode: Executes the full scoped task until completion or a genuine blocker is reached, without unnecessary interactive pauses.
  • Clarification protocol and completion contract: Records ambiguity in the completion report’s Blockers field using CLARIFICATION_NEEDED tags and returns a minimum report with Status, Summary, and Blockers so the caller can proceed.

Use case: when mission-control delegates a bounded step to another agent, this pattern ensures the sub-agent runs to completion, and if requirements conflict it requests resolution through a structured CLARIFICATION_NEEDED blocker rather than turning it into a direct user question.

Quick Start

Ask your sub-agent to follow the handoff-mode rules and end every execution with a minimum completion report containing Status, Summary, and Blockers.

Frequently Asked Questions about sub-agent-patterns

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

FAQPage Schema
How do I handle agent handoff when an orchestrator delegates tasks to sub-agents?▼

Agent handoff requires structured delegation using Task, Context, Constraints, and Success criteria. This pattern ensures sub-agents execute autonomously to completion or stop with actionable blocker metadata, enabling safe orchestration workflows without unnecessary interactive pauses.

How does an autonomous execution gating protocol work for AI agents?▼

Autonomous execution gating allows a sub-agent to run the full scoped task until completion or a genuine blocker is reached. The protocol prevents unnecessary interactive pauses by enforcing mission-control-style delegation cues and requiring a structured completion report before returning control.

What is a CLARIFICATION_NEEDED protocol for AI agent orchestration?▼

The CLARIFICATION_NEEDED protocol records ambiguity in a completion report's Blockers field. Instead of asking the user directly, sub-agents tag conflicting requirements as CLARIFICATION_NEEDED so the orchestrating agent can resolve them structurally.

What should a minimum completion report contract include for agent-to-agent handoffs?▼

A minimum completion report contract must include Status, Summary, and Blockers fields. This structured format ensures the calling orchestrator can safely continue orchestration by understanding the execution outcome and any actionable blockers encountered.

How do I structure a prompt for safe sub-agent delegation in autonomous workflows?▼

Structure sub-agent delegation prompts using Task, Context, Constraints, and Success criteria. This mission-control-style format enables handoff-mode inference and ensures the sub-agent executes the bounded step safely to completion.

Why does my sub-agent stop and ask questions during autonomous task execution?▼

Sub-agents stop during autonomous execution when they encounter conflicting requirements. Instead of pausing interactively, this pattern ensures they use the CLARIFICATION_NEEDED tag within the completion report's Blockers field to request resolution structurally.

Related Skills

Explore More Software Engineering Skills→
pyronaurpyronaur

prompting

Design prompt contracts with inputs, outputs, constraints, and success criteria.

Community
Advanced
semicolon-devteamsemicolon-devteam

generate-member-report

Generate per-member workload and status reports with completion rates and blockers.

Official
Intermediate
arbazkhan971arbazkhan971
26

agent

Design and validate autonomous AI agents with tool orchestration and guardrails.

Community
Advanced
Leonardo-CorteLeonardo-Corte
1

maestro:conduct

Orchestrate mission execution by decomposing tasks, dispatching sub-agents, and synthesizing results.

Community
Advanced