subagents-discipline

Guide subagents to implement feature changes with context inspection and functional verification.

Updated Feb 10, 2026
One-click install
npx skills add https://github.com/jordanhindo/meridian --skill subagents-discipline-jordanhindo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagents-discipline
Source: https://github.com/jordanhindo/meridian/tree/main/.claude/skills/create-beads-orchestration/templates/skills/subagents-discipline
Command: npx skills add https://github.com/jordanhindo/meridian --skill subagents-discipline-jordanhindo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents implementation mistakes by enforcing a disciplined workflow for reading the provided context, inspecting real data before coding, and verifying behavior quickly.

Core Features & Use Cases

  • Bead-first execution: Ensures you use existing dispatch and root-cause context so you don’t re-investigate or miss edge-case findings.
  • Data-then-code accuracy: Requires fetching the actual external-data shape (fields, formats, types) before implementing queries or integrations.
  • Close-the-loop verification: Promotes fastest functional checks (curl/CLI/browser/database inspection) over “tests only” verification.
  • Tool-aware validation: Encourages using available tools (e.g., MCP/browser/API inspection) when full testing seems blocked.
  • Optional reasoning log: Supports maintainability by documenting deviations when choices might be questioned later.
  • Epic contract adherence: Guides subagent work to align exactly with epic design docs when implementing children of larger features.

Quick Start

Ask the orchestrator for the bead’s dispatch context if no dispatch or context comments exist, then follow the rules to implement and verify the change using the logged findings.

Frequently Asked Questions about subagents-discipline

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

FAQPage Schema
How do I prevent implementation mistakes when integrating external APIs or databases?

Prevent implementation mistakes by fetching the actual external-data shape, verifying field names and types before coding, and closing the loop with fast functional checks like curl or CLI inspection.

What is the best way to manage engineering workflow for subagent feature changes?

The best way to manage subagent feature changes is enforcing bead-first execution, which uses existing dispatch and root-cause context to avoid re-investigation and ensure edge-case findings are applied.

How do I verify feature changes when automated tests are blocked?

Verify feature changes using tool-aware validation, which encourages using available tools like MCP, browser, or API inspection to perform fastest practical functional checks instead of relying only on tests.

Does this engineering workflow approach work for implementing epic child features?

Yes, this workflow approach works for epic child implementations by enforcing epic contract adherence, guiding subagent work to align exactly with the parent epic's design documents during implementation.

When should I document deviations in my engineering workflow?

Document deviations in your engineering workflow when implementation choices might be questioned later, using the optional reasoning log to support maintainability and provide context for future maintainers.