What is the Orchestrator's role vs. the Teammate's role?

The Orchestrator manages the overall process, routes work, and makes high-level judgment calls, while Teammates supervise specific implementation tasks and manage subagent review loops.

How does it ensure code quality during implementation?

The skill enforces a 'Control Contract' that requires evidence references for all commits and explicit dispositions for all findings before a story can be accepted.

Does this skill require specific models to function?

It is optimized for a 'Codex lane' using GPT-5.3-codex for implementation, but includes a Sonnet-only fallback mode if specialized subagent skills are not available.

Can this skill handle project-specific testing requirements?

Yes, it performs a 'Verification Gate Discovery' phase on load to identify required unit, integration, and E2E commands from your project configuration.

What is the purpose of the team-impl-log.md file?

It is a persistent orchestration log that tracks lane decisions, verification gate results, process adaptations, and story-level observations throughout the implementation cycle.

Team Implementation Orchestrator

Name: Team Implementation Orchestrator
Author: liminal-ai

byliminal-ai

0•

コラボレーションツール

Orchestrates multi-agent teams to execute complex software implementations through sequential story management and rigorous verification cycles.

The ls-team-impl skill transforms Claude into a high-level project orchestrator capable of managing specialized agent teams for large-scale development tasks. It coordinates the implementation of technically enriched stories and epics by spawning teammates, managing handoffs, and supervising Codex subagents. By enforcing a strict 'Control Contract' and maintaining a persistent orchestration log, the skill ensures that every code change is verified against project-specific gates, reducing the need for human intervention while maintaining high architectural standards and spec compliance.

主な機能

01Persistent orchestration logging for state tracking across complex epics

020 GitHub stars

03Automated discovery and enforcement of project-specific verification gates

04Dual-lane execution support for specialized Codex models or Sonnet fallbacks

05Multi-round self-review loops and cross-model verification patterns

06Autonomous multi-agent team orchestration and teammate lifecycle management

ユースケース

01Executing multi-story epics from technical designs with high architectural fidelity

02Managing complex code migrations that require sequential, verified implementation steps

03Automating the build-test-review cycle for large-scale features using specialized subagents

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add liminal-ai/code-steward ls-team-impl

For use in Claude.ai and ChatGPT

主な機能

01Persistent orchestration logging for state tracking across complex epics

020 GitHub stars

03Automated discovery and enforcement of project-specific verification gates

04Dual-lane execution support for specialized Codex models or Sonnet fallbacks

05Multi-round self-review loops and cross-model verification patterns

06Autonomous multi-agent team orchestration and teammate lifecycle management

ユースケース

01Executing multi-story epics from technical designs with high architectural fidelity

02Managing complex code migrations that require sequential, verified implementation steps

03Automating the build-test-review cycle for large-scale features using specialized subagents