skills.rest
e-t-y-b
e-t-y-bOFFICIAL
Explore all skills by this author →
GitHub

plan-execution-protocol

Coordinates task-by-task execution with verification and gate transitions.

9|1|Updated Mar 21, 2026
Search Tags:#subagent#plan-execution#blocker-management#task-verification#inline-execution#gate-transitions#tdd-protocol
One-click install
npx skills add https://github.com/e-t-y-b/etyb-skills --skill plan-execution-protocol
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: plan-execution-protocol
Source: https://github.com/e-t-y-b/etyb-skills/tree/main/skills/plan-execution-protocol
Command: npx skills add https://github.com/e-t-y-b/etyb-skills --skill plan-execution-protocol

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Plan execution runtime — drives task-by-task implementation with per-task verification, blocker management, and gate transitions. One task at a time, verified before advancing, plan updated after every task. Use when executing, resuming, or checking status of an active plan.

Core Features & Use Cases

  • Plan execution loop and deterministic task handling across gates
  • Per-task verification, blocker management, and gate transitions
  • One-task-at-a-time discipline with plan mutation and clear handoffs
  • Use when executing, resuming, or auditing an active engineering plan

Quick Start

Initiate a run by loading an active plan and selecting the next pending task for execution.

Frequently Asked Questions about plan-execution-protocol

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

FAQPage Schema
How do I execute an engineering plan one task at a time with verification?▼

Plan execution runtime drives task-by-task implementation by loading plan artifacts, executing the next pending task in inline, subagent, or hybrid mode, and verifying results before advancing to the next gate.

What is the best way to manage blockers during plan execution?▼

Blocker management is handled natively by the plan execution runtime, which captures blockers encountered during task execution, pauses the current workflow gate, and updates the plan artifact to reflect the blocked state.

How does gate transition work when executing a multi-step plan?▼

Gate transitions occur deterministically after per-task verification succeeds, allowing the runtime to commit verified results, mutate the active plan to mark the task complete, and advance to the next pending task in the workflow.

Can I resume an active plan from the last pending task after a interruption?▼

Yes, plan execution supports resuming active plans by loading the existing plan artifact, identifying the next pending task, and continuing the execution loop without re-running previously verified and committed tasks.

Does plan execution support subagent mode for task implementation?▼

Yes, the runtime supports inline, subagent, and hybrid execution modes, allowing you to load expert skills and delegate individual task implementation to subagents before verifying results and committing them to the plan.

When should I not use a one-task-at-a-time execution protocol?▼

One-task-at-a-time execution is not suited for parallel task workflows, as its core discipline enforces sequential, verified gate transitions and deterministic plan mutation across multi-gate engineering workflows.

Related Skills

Explore More Software Engineering Skills→
fitlab-aifitlab-ai
84

plan-task

Generate technical design plans and implementation roadmaps for human review.

Official
Advanced
rockandrolla13rockandrolla13

plan-tracker

Create, update, and verify PLAN-*.md files for multi-step workflows.

Community
Advanced
KartikSharma4448KartikSharma4448

delegation

Delegate tasks to subagents for synchronous or background execution.

Community
Advanced
boneskullboneskull
8

executing-plans

Execute implementation plans in batches with review checkpoints.

Community
Basic