Claude-skill-registry implementation-guidance

Provides detailed guidance on how to implement a single plan step while adhering to architecture and quality standards.

install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/implementation-guidance" ~/.claude/skills/majiayu000-claude-skill-registry-implementation-guidance && rm -rf "$T"
manifest: skills/data/implementation-guidance/SKILL.md
source content

Implementation Guidance Skill

This skill helps you turn a single plan step into code changes while following TDD and maintaining architectural integrity.

Steps

  1. Select a plan slice. Identify the next unimplemented milestone or task from

    PLAN.md
    . Summarise its objective, inputs, outputs and acceptance criteria.

  2. Review constraints. Load

    ARCHITECTURE.md
    and
    CONTRIBUTING.md
    to understand any constraints (e.g. design patterns, layer boundaries, security policies). Ensure your implementation will not violate them.

  3. Write tests first. Before writing production code, design unit and integration tests that express the desired behaviour. Use the testing strategy guidelines and harness appropriate frameworks.

  4. Implement incrementally. Write small, focused commits that satisfy one test at a time. Avoid large diffs. Document significant decisions or deviations in an ADR if necessary.

  5. Run validation and linting. After coding, execute the validation task (

    Context Kit: Validate
    ) and any language-specific linters or static analysis tools. Fix issues immediately.

  6. Summarise changes. Prepare a concise summary of what was changed, which files were touched and how the acceptance criteria were met. Provide links to relevant docs or ADRs.

  7. Handoff for review. Once the step is complete and validated, hand off to the reviewer agent or trigger the

    review-changes
    prompt for quality assurance.

By following this process you produce maintainable code that is easy to review and less likely to introduce regressions or technical debt.