converge

Orchestrate end-to-end plan convergence from tickets, threads, and drafts into structured work items.

1|Updated May 15, 2026
One-click install
npx skills add https://github.com/mslshao/claude-harness --skill converge-mslshao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: converge
Source: https://github.com/mslshao/claude-harness/tree/main/dotclaude/skills/converge
Command: npx skills add https://github.com/mslshao/claude-harness --skill converge-mslshao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes the manual overhead of turning scattered product or engineering inputs into a trustworthy execution plan. It helps when you have a rough idea, ticket, thread, or draft and want a production-quality plan that has already been challenged, reviewed, and decision-gated before anyone starts implementation.

Core Features & Use Cases

  • End-to-end planning orchestration: Runs refinement, decomposition, challenge, specialist consultation, synthesis, skeptic review, and a proceed-or-iterate gate as one workflow.
  • Multi-source input handling: Accepts free text, Jira tickets, bead references, Slack threads, Confluence pages, PRs, and transcripts, then reconciles conflicts with explicit precedence rules.
  • Mechanism-bias detection: Detects when a user has prescribed a solution too early and forces the workflow to test whether that mechanism is actually the right one.
  • High-confidence planning output: Produces structured work items with dependencies, verification paths, consequence-of-wrong, convergence delta, and open assumptions.
  • Use case: When a team has a loosely defined feature request spread across Jira, Slack, and prior notes, use this Skill to produce a converged implementation plan without manually chaining multiple planning and review skills.

Quick Start

Ask the AI to use the converge skill on your rough feature idea, Jira ticket, Slack thread, or draft doc to produce a converged plan with vetted work items and verification paths.

Frequently Asked Questions about converge

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

FAQPage Schema
How do I turn rough ideas and Slack threads into a production-ready implementation plan?

Implementation plan convergence orchestrates refinement, decomposition, and skeptic review to turn rough ideas, Slack threads, and tickets into vetted work items with dependency mapping and verification paths.

What is the best way to create a technical design from scattered Jira tickets and draft docs?

Technical design from scattered Jira tickets and drafts is generated through multi-source input handling, which reconciles conflicts using explicit precedence rules before producing structured work items.

Can I use workflow orchestration to detect when a solution mechanism is prescribed too early?

Yes, workflow orchestration applies mechanism-bias detection to identify when a solution mechanism is prescribed too early, testing whether that mechanism is actually the correct choice for the plan.

Does planning orchestration require code generation to map dependencies and verify work items?

No, planning orchestration does not require code generation; it enforces verification paths, convergence delta tracking, and structured work-item output without writing code.

How do I proceed-or-iterate gate a roadmapping workflow before execution?

To proceed-or-iterate gate a roadmapping workflow, the orchestration runs an adversarial review and mandatory skeptic review to ensure the plan is challenged and decision-gated before execution.

When should I not use automated plan convergence for technical design?

You should not use automated plan convergence when your technical design lacks multi-input sources like transcripts or threads, as the workflow relies on reconciling scattered inputs to produce trustworthy work items.