wave-executor

Coordinate multi-agent engineering sessions into structured waves with quality checks.

50|7|Updated Apr 2, 2026
One-click install
npx skills add https://github.com/Kanevry/session-orchestrator --skill wave-executor-kanevry
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wave-executor
Source: https://github.com/Kanevry/session-orchestrator/tree/main/skills/wave-executor
Command: npx skills add https://github.com/Kanevry/session-orchestrator --skill wave-executor-kanevry

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill coordinates complex engineering sessions by executing an agreed plan in controlled waves, preventing scope creep, uncoordinated commits, and cascading failures across parallel agents.

Core Features & Use Cases

  • Wave-based orchestration: Dispatches role-mapped agents in parallel, waits for completion, reviews outputs, adapts the plan, and advances wave-by-wave.
  • Safety & recovery: Enforces turn limits, detects spirals and stagnation patterns, applies circuit-breaker recovery, and isolates agent worktrees when configured.
  • Quality gating & metrics: Runs incremental and full quality checks, session-reviewer analyses, grounding injection, and persistent session metrics for auditing and reproducibility.
  • Use Case: Execute a feature or deep engineering session with multiple implementers and reviewers while preserving reproducible state, tracked progress, and robust failure handling.

Quick Start

Invoke the coordinator by running /go to execute the agreed session plan in waves while enforcing scope, quality gates, and progress reporting.

Frequently Asked Questions about wave-executor

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

FAQPage Schema
How do I orchestrate multi-agent execution for parallel engineering tasks?

Multi-agent execution is orchestrated by dispatching role-mapped agents in structured waves, coordinating parallel work, reviewing outputs, adapting plans, and advancing wave-by-wave to prevent scope creep and uncoordinated commits.

What is the best way to prevent cascading failures across parallel agents?

Preventing cascading failures across parallel agents requires enforcing turn limits, detecting spirals and stagnation patterns, applying circuit-breaker recovery, and isolating agent worktrees to contain errors during multi-agent execution.

How do I enforce quality gates during a multi-agent engineering session?

Quality gates are enforced by running incremental and full quality checks, session-reviewer analyses, and grounding injection between execution waves, ensuring outputs meet standards before advancing the session.

Can I recover and reproduce state from a failed multi-agent engineering session?

Yes, sessions are made recoverable and reproducible through persistent session metrics, worktree isolation, and circuit-breaker detection, allowing you to audit and resume failed multi-agent runs without losing progress.

Does worktree isolation work with role-based parallel agents?

Yes, worktree isolation is applied to role-based parallel agents, preventing uncoordinated commits and managing conflict resolution when multiple agents execute tasks within the same engineering session.

When should I use wave-based orchestration for engineering sessions?

Wave-based orchestration should be used for feature, deep, and housekeeping engineering sessions that require role-based parallel agents, inter-wave reviews, scope manifests, and robust failure handling to prevent scope creep.