orchestrate-dev

Runs a full development pipeline covering architecture, planning, implementation, and iterative code review.

1|1|Updated Feb 25, 2026
One-click install
npx skills add https://github.com/williamthorsen/codeassembly --skill orchestrate-dev-williamthorsen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrate-dev
Source: https://github.com/williamthorsen/codeassembly/tree/main/packages/agents/content/skills/orchestrate-dev
Command: npx skills add https://github.com/williamthorsen/codeassembly --skill orchestrate-dev-williamthorsen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Coordinating architecture assessment, planning, coding, and multi-round code review by hand is error-prone and inconsistent. This Skill orchestrates the entire development workflow through a single engine invocation with configurable effort levels. ## Core Features & Use Cases - Effort presets: Choose low, medium, or high effort to control review rounds, approval thresholds, and aspect reviewers, with per-setting CLI overrides. - Configurable pipeline: Architecture and planning phases are optional and can be required or skipped via flags; implementation and review-cycle always run. - Findings-based review handling: F/W/T/R/S findings are fixed, deferred, ticketed, or piggybacked according to the effort level and thresholds. - Use Case: Ask the agent to implement a new API endpoint at high effort; it plans the change, writes the code, runs parallel and holistic reviews across up to four rounds, and tickets deferred TODOs at wrap-up. ## Quick Start Ask the agent to implement your task with the orchestrate-dev workflow, for example: run the full development workflow to add user authentication with high effort.

Frequently Asked Questions about orchestrate-dev

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

FAQPage Schema
How do I run a full development workflow with code review?

Invoke the orchestrate-dev skill with a task description; it runs architecture, planning, implementation, and review-cycle phases through the orchestrate engine. Use --effort to control how much review work is performed.

What do the low, medium, and high effort levels change?

Effort presets adjust approval-threshold, budget-threshold, max review rounds (2, 3, or 4), and aspect reviewers. Higher effort fixes more finding tiers and spends remaining review budget on lower-severity findings.

Can I override individual settings instead of using an effort preset?

Yes. Explicit CLI arguments like --approval-threshold or --max-review-rounds take highest priority, followed by the effort preset, then preferences.yaml orchestration keys, then engine defaults.

Is the --mode argument still supported?

The --mode argument is deprecated in favor of --effort. If only --mode is given, vibe maps to low, lite to medium, and strict to high; when both are present, --effort takes precedence.

What happens to findings that are not fixed during the run?

Deferred TODO findings become tracked tickets at wrap-up, deferred Recommendations are noted in the run summary, and Suggestions are not tracked. The wrap-up skill is offered automatically when deferred items exist.