Evaluate design options and write the chosen approach to tasks/design-decision.md.
npx skills add https://github.com/KMakayee/playbook --skill design-kmakayee
Please help me install this Agent Skill. Skill: design Source: https://github.com/KMakayee/playbook/tree/main/.claude/skills/design Command: npx skills add https://github.com/KMakayee/playbook --skill design-kmakayee
It prevents flawed architecture decisions by forcing a research-grounded design proposal and validating the choice through an independent cross-check.
tasks/design-decision.md with options, evaluation criteria, and a clear chosen approach ready for planning.Tell the AI to run the design workflow after you have created the research artifact by executing /research-codebase and then executing /design.
High-intent search queries and answers about installing and using this skill.
To choose a software design, you must reconcile research evidence, constraints, and risk trade-offs across multiple viable options. This Skill evaluates design alternatives and uses an independent Codex cross-check to validate the chosen approach before finalizing the decision.
An independent design cross-check runs a separate generation process to evaluate your proposed architecture options. It surfaces missed alternatives and risk trade-offs, verifying that your chosen approach is justified and implementation-ready before you commit.
You document an architecture decision by writing the evaluation criteria, viable options, and final chosen approach to a design-decision.md file. This artifact ensures the implementation planning phase has a verified, actionable decision record with resolved blocking questions.
Yes, prerequisite codebase research is required before evaluating design options. This Skill verifies prerequisite research artifacts exist and uses them to generate viable architecture options, ensuring the final decision is grounded in actual codebase evidence.
Research alone for software planning can miss alternatives and risk trade-offs. This Skill mitigates the limitation by running an independent cross-check against your research-grounded options, verifying claims and resolving blocking questions before the decision is finalized.
xiechyGenerate testable hypotheses with mechanistic explanations and experimental designs.
ricardoalt1515Generate structured task breakdowns for SDD changes using Python and Markdown.
drakulavichArchive completed OpenSpec changes after validating artifact status and delta specs.
juanmanueldazaGenerate and evaluate 2–3 approaches for planned software changes.