Implement

This is a heavyweight orchestrator that spins up parallel subagents to build features across backend, frontend, testing, and security in one coordinated run. It handles full-stack implementations with worktree isolation, automatic tier classification from tier 1 (Interview Starter) to tier 6 (Open Source), and budget-aware phase scaling based on your effort setting. The thing that stands out is the resume capability: it writes state checkpoints so you can stop mid-implementation and pick up exactly where you left off. Designed for when you're building something substantial and want coordinated architecture, code generation, and test coverage without manually stitching together multiple agent sessions. Works best with the memory and context7 MCP servers loaded, though it will adapt if they're missing.

skills.sh installs: 206
GitHub stars: 301
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add yonatangross/orchestkit --skill implement --agent claude-code

Installs into .claude/skills of the current project.

Files

Implement Feature

Host-neutral workflow. Invoke by skill name (implement). Claude Code slash routing, YAML hook loaders, and .claude/chain live in references/claude-code.md.

Parallel subagent execution for feature implementation with scope control and reflection.

Quick Start

implement user authentication
implement --model=opus real-time notifications
implement dashboard analytics

Argument Resolution

Check the target first, before any other step. This skill forks and cannot see the conversation. Load Read("../../shared/rules/target-resolution.md"): if the target, flags stripped, is a bare pronoun or deictic (this, them, that one) or empty, return exactly TARGET_UNRESOLVED: <word> and stop. The caller resolves it and re-invokes.

FEATURE_DESC = "$ARGUMENTS"  # Full argument string, e.g., "user authentication"
# $ARGUMENTS[0] is the first token, $ARGUMENTS[1] second, etc. (CC 2.1.59)

SKILL.md on GitHub