implementation

The Implementation skill provides a robust, project-agnostic framework for managing software feature development within the IMPLEMENT state of an agentic workflow. It acts as a systematic orchestrator that ensures development follows strict quality assurance cycles, preventing premature commits and ensuring that every feature undergoes rigorous verification. By enforcing a template-based structure via .claude/scripts/, the skill allows developers to tailor commands—such as health checks, commit conventions, and build processes—to specific project needs, such as TypeScript/Node, Python, Go, or Rust environments.

• End-to-end management of feature lifecycles using feature-list.json as the source of truth.
• Critical gatekeeping through TEMPLATE-*.sh files that must be customized to prevent inconsistent deployment and development patterns.
• Systematic sub-workflow for feature implementation: code generation, unit testing, type checking, project building, and health status verification.
• Automated session introspection that audits command history to detect friction points and suggests script updates for continuous workflow optimization.
• Token-efficient codebase analysis using dedicated MCP tools for efficient log parsing, code execution, and large dataset handling.
• Integration with context-graph systems for storing architectural decisions and project learnings to maintain long-term consistency.

When using this skill, users should ensure the .claude/scripts/ directory is fully configured. The workflow is designed to be blocking: if templates are present, the agent will refuse to proceed until they are renamed and verified. Always ensure your environment supports the necessary verification commands like pnpm typecheck, mypy, or cargo check as dictated by your specific tech stack. The skill assumes you are following a commit-only-after-success policy, ensuring that the main branch remains stable and that manual testing is performed whenever automated suites might be insufficient to confirm functionality. It is ideal for developers managing complex multi-session feature sets who want to maintain high code standards and reduce technical debt through automated feedback loops.