Plan, implement, and verify multi-step coding tasks with user approval.
npx skills add https://github.com/LudwigAJ/swell --skill code-ludwigaj
Please help me install this Agent Skill. Skill: Code Source: https://github.com/LudwigAJ/swell/tree/main/.claude/skills/code Command: npx skills add https://github.com/LudwigAJ/swell --skill code-ludwigaj
This Skill helps turn coding requests into a clear, safe workflow so software changes are planned before they are built, verified before they are shared, and aligned with user preferences.
Use the Code skill to plan the requested change, implement it in your project, and verify the result with the appropriate tests before reporting back.
High-intent search queries and answers about installing and using this skill.
Planning multi-step coding tasks involves breaking implementation work into testable steps with clear progress checkpoints. This structured workflow requires explicit user approval between steps to ensure changes align with preferences before any execution begins.
The best way to verify code changes is through a verification-first approach that recommends writing tests, capturing screenshots, or performing output checks before delivery to ensure software changes are fully validated and safe.
Yes, a structured workflow applies to feature implementation, bug fixing, refactoring, and other multi-step development tasks. It guides software implementation from planning through verification to ensure safe, careful execution.
No, this coding workflow does not execute changes autonomously. It requires explicit user approval between steps and provides preference-aware guidance to ensure careful step-by-step execution instead of rushed changes.
User coding preferences are stored in a local memory file. The system saves only user-approved coding preferences locally to align future guidance and implementation steps with your specific workflow standards.
antgroupPlan, implement, verify, and test code with file-based memory storage.
dfrysingerRun acceptance tests against implemented code and present pass/fail results.
semicolon-devteamAutomate Playwright end-to-end tests to validate runtime browser behavior.
haJ1tConstruct self-contained prompts for AI agents to execute multi-step tasks.
