do

Orchestrates subagents to execute phased implementation plans with verification checkpoints.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills --skill do-lgj-jonathan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: do
Source: https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills/tree/main/claude-mem/plugin/skills/do
Command: npx skills add https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills --skill do-lgj-jonathan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Executing a multi-phase implementation plan manually is error-prone: steps get skipped, unverified code gets committed, and APIs get invented instead of checked against documentation. This Skill turns plan execution into a disciplined orchestration process where every phase is implemented, verified, and reviewed by dedicated subagents before any commit happens. ## Core Features & Use Cases - Subagent Orchestration: Deploys fresh Implementation, Verification, Anti-pattern, Code Quality, Commit, and Branch/Sync subagents for each phase of a plan. - Verification Gates: Blocks commits until a verification checklist passes, and greps for known anti-patterns from the plan after every phase. - Documentation-Grounded Implementation: Requires subagents to copy patterns from documentation, cite sources in code comments, and stop when an API cannot be verified. - Use Case: After generating a phased plan with a planning skill, hand it off here so each phase is executed by fresh subagents, verified against a checklist, reviewed for code quality, and committed to a working branch only after passing. ## Quick Start Execute the implementation plan from make-plan phase by phase, verifying and committing each phase before moving on.

Frequently Asked Questions about do

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

FAQPage Schema
How do I execute a multi-phase implementation plan with AI subagents?

Deploy one Implementation subagent per phase with a single clear objective, then run Verification, Anti-pattern, and Code Quality subagents before committing. Require each subagent to report evidence such as commands run, outputs, and files changed.

How to prevent AI agents from inventing APIs during implementation?

Instruct subagents to copy patterns directly from documentation, cite documentation sources in code comments, and stop to verify when an API seems missing. Never allow undocumented parameters or assumed signatures.

Should code be committed before verification passes?

No. Commits should only happen after a verification subagent runs the phase checklist and confirms the work matches the plan. A Commit subagent is deployed only after verification passes or with explicit orchestrator approval.

What is the difference between planning and executing a plan with AI?

Planning produces the phased checklist and context, while execution orchestrates subagents to carry out each phase. This Skill handles only execution and works best on plans created by a separate planning step such as make-plan.

When should fresh subagents be used between phases?

Use fresh subagents when the plan notes it or when context becomes large or unclear. A Branch/Sync subagent pushes the verified work and prepares a handoff so the next phase starts clean but retains plan context.