sp-implement-teams

Execute spec-kit implementation tasks through Claude Code Agent Teams with shared task ledgers.

Updated Jul 6, 2026
One-click install
npx skills add https://github.com/chenziyang110/launchdeck --skill sp-implement-teams-chenziyang110
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sp-implement-teams
Source: https://github.com/chenziyang110/launchdeck/tree/main/.claude/skills/sp-implement-teams
Command: npx skills add https://github.com/chenziyang110/launchdeck --skill sp-implement-teams-chenziyang110

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Coordinating multi-step feature implementation across parallel workers often loses task state, verification evidence, and completion handoffs. This Skill runs the spec-kit implementation phase through Claude Code Agent Teams so every task carries an explicit execution packet, dependency wiring, and structured result handoff. ## Core Features & Use Cases - Team Bootstrap Gate: Creates or resumes a Claude Agent Team per feature slug before any implementation context is read, preventing duplicate teams and silent fallbacks to ordinary subagents. - Shared Task Ledger: Converts ready tasks from tasks.md into shared tasks with blockedBy/blocks dependencies, owner assignment, write sets, verification commands, and result handoff paths. - Wave Protocol & Result Contract: Coordinates implementation, review, and verification teammates with structured messages (task_started, task_blocked, task_completed) and normalizes WorkerTaskResult handoffs before declaring completion. - Use Case: After running /sp-tasks for a feature, invoke /sp-implement-teams to have a leader agent dispatch implementation slices to durable teammates, track progress in ~/.claude/tasks/, and collect verified results per task. ## Quick Start Ask the agent to run /sp-implement-teams for the current feature so it bootstraps a Claude Agent Team and executes the ready tasks from tasks.md with shared tracking.

Frequently Asked Questions about sp-implement-teams

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run spec-kit implementation with Claude Agent Teams?▼

Invoke /sp-implement-teams after tasks.md exists for the feature. The skill runs check-prerequisites, creates or resumes a Claude Agent Team for the feature slug, converts ready tasks into shared tasks with dependencies, and launches teammates that report structured handoffs.

What is the difference between /sp-implement and /sp-implement-teams?▼

Both share the same execution contract, tracker semantics, and completion discipline. /sp-implement may route work through ordinary subagents, while /sp-implement-teams forces team-managed execution through Claude Code Agent Teams with a shared ledger and explicit teammate messaging.

Does sp-implement-teams work without Claude Code Agent Teams enabled?▼

No. Agent Teams availability is a hard prerequisite checked via the CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS setting or environment. If TeamCreate or the native teammate launch surface is unavailable, the skill stops and reports the missing feature instead of falling back to ordinary subagents.

Why does the skill refuse to use the ordinary Agent tool for teammates?▼

The workflow requires team-managed teammates that join the shared Agent Teams ledger under ~/.claude/teams/ and ~/.claude/tasks/. The ordinary Agent tool cannot join that ledger, so using it would break dependency tracking, ownership, and completion handoff guarantees.

How are teammate results validated before completion is declared?▼

Teammates must send task_started, task_blocked, and task_completed messages plus write a WorkerTaskResult file to FEATURE_DIR/worker-results/<task-id>.json when a handoff is promised. The leader normalizes statuses like DONE_WITH_CONCERNS or NEEDS_CONTEXT and never treats an idle teammate as finished work.