careful-impl

Execute post-plan-mode plans with self-review and task decomposition.

Updated Nov 12, 2017
One-click install
npx skills add https://github.com/ak110/dotfiles --skill careful-impl
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful-impl
Source: https://github.com/ak110/dotfiles/tree/main/plugins/agent-toolkit/skills/careful-impl
Command: npx skills add https://github.com/ak110/dotfiles --skill careful-impl

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill executes post-plan-mode plans with a self-review step, ensuring a structured workflow from decomposition to commit after plan agreement.

Core Features & Use Cases

  • Task decomposition and registration using TaskCreate, deriving dependencies and sequencing validation, review, and commit tasks.
  • Delegation of concrete work to the careful-implementer, while loading coding- and writing-standards automatically from the frontmatter.
  • Parallelized review by spec, code, and docs reviewers to ensure multi-dimensional quality before final commits.
  • Clear handoff and traceability for tracking plan execution progress across sub-agents and workflows.

Quick Start

Provide the plan file to this skill and let it orchestrate decomposition, implementation, validation, review, and commit steps.

Frequently Asked Questions about careful-impl

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

FAQPage Schema
How do I automate plan execution with a self-review step after plan approval?

Automated plan execution with self-review takes an approved plan file, decomposes it into implementation, validation, and commit phases, then delegates work to subordinate agents. It loads coding standards from frontmatter to guide agents and ensure traceability.

What is task decomposition in multi-phase software planning workflows?

Task decomposition in multi-phase software planning uses TaskCreate to derive dependencies and sequence validation, review, and commit tasks. It breaks an approved plan into discrete phases managed by subordinate agents for structured execution.

How do I parallelize code, spec, and docs reviews before a final commit?

Parallelized review of spec, code, and docs assigns each dimension to a separate reviewer before final commits. This ensures multi-dimensional quality by evaluating specifications, implementation, and documentation concurrently.

Do I need a plan file to start task decomposition and validation workflows?

Yes, a plan file is the sole input required to start task decomposition and validation workflows. The skill reads frontmatter-loaded standards from this file to guide subordinate agents through implementation and review.

What's the best way to track execution progress across sub-agents in a software workflow?

Tracking execution progress across sub-agents relies on clear handoff and traceability built into the workflow. Task decomposition registers dependencies via TaskCreate, allowing you to monitor implementation, validation, and review stages.

When should I not use automated self-review for plan execution?

Automated self-review for plan execution requires an approved plan file with frontmatter standards. If your workflow lacks multi-phase decomposition needs or a structured plan file, manual implementation and review may be more appropriate.