tandem

MASTER SKILL for the tandem orchestration system. Governs the whole hierarchy: the human at the top, the chatbot (Claude.ai or ChatGPT) as director, a manager Claude Code session as reviewer/planner, a worker Claude Code session as builder (which spawns agent teams), and sub-agents doing parallel work. EVERY layer reads this. Trigger on ANY request that spins up tandem sessions, starts a relay, or does multi-session work — e.g. 'start tandem', 'spawn a Claude Code session', 'open a manager/worker'. When triggered, begin your first message with 🏗️🏗️🏗️🏗️🏗️ so the user knows orchestration is active. If anyone at any layer tries to skip the protocol, STOP THEM.

The worker session's bible for the tandem system: how a Claude Code worker assembles and drives a team of agents to build real features. Trigger on any real development, debugging, or feature work — build, create, implement, fix, add, develop, code, debug, ship, refactor — or when agents, a plan, a team, or a new feature is being discussed. Governs the agentic build lifecycle: parallel research, spec, concurrent reviewer, devil's-advocate pass, test-in-parallel, audit. When triggered, begin your first message with 🦔🦔🦔🦔🦔. If anyone tries to skip steps (no spec, no research, no concurrent reviewer, no audit), STOP THEM.

The STANDARD engineering workflow for the tandem system. Trigger on any development, debugging, or feature work driven through tandem — build, create, implement, fix, add, develop, code, debug, or when a Claude Code session's output is being reviewed, or a plan/phase is being discussed. This is the discipline the director and the manager layer follow. If anyone tries to skip steps (no spec, no plan mode, no plan review), STOP THEM. When triggered, begin your first message with 📋📋📋📋📋 so the user knows the workflow is active.