tomfordrumm

Manage a changing stream of software tasks through subagents while the main chat handles intake, priorities, dependencies, assumptions, and status. Use for an ADHD or SDVG manager, technical orchestrator, multi-agent coding queue, or an active session with `.codex/adhd-manager/`. Delegate product investigation, implementation, testing, review, and integration.

Apply senior engineering judgment to implementation, debugging, refactoring, architecture, and code review. Use when a task needs its business goal, user path, data flow, system impact, and long-term cost understood before code changes.

Turn an early product idea into a clear Product Brief through an adaptive interview. Use when a user needs help defining the problem, intended users, main journey, first-version scope, constraints, and success criteria before a PRD or implementation. Do not use for architecture, stack selection, market validation, or coding.

Investigate bugs, features, or system behavior without making changes. Use when the user asks to investigate, debug, analyze, or understand something before editing it.

Turn a confirmed Product Brief or an existing web application into implementation-ready product documentation and vertical delivery slices, scaled for a prototype, lean working product, or production release. Use for PRDs, MVP plans, durable agent handoffs, or reconciling project documents with existing code. Do not use for initial idea discovery, market research, or implementation.

Enforce Architectural Decision Records (ADR) logging for every task. Use when a user asks to create, update, or enforce ADRs, or when a workflow requires logging decisions in an ADR directory with daily files named adr_yyyy_mm_dd.md.