dev

Use only when the user explicitly asks or allows delegating implementation to a Codex worker — launch Pi running openai-codex/gpt-5.6-sol (xhigh) as a visible sidecar implementer that this session orchestrates over AMQ. Write exact task specs, dispatch with --kind todo, validate the results yourself, then iterate, resume, respawn, or launch another worker. Never invoke on your own initiative.

Use only when the user explicitly allows or asks for this skill to be invoked to pair with a second agent harness — launch a sidecar, get a second opinion from another agent, have one harness drive another, or review work with another agent over AMQ. Otherwise, never invoke it.

first step of planning

convert structure outline into a detailed implementation plan

Create a Product Requirements Document with inline mockups

generate research questions based on a task, spec, or ticket

research the codebase

create a phased implementation plan based on research and design decisions

Create a Technical Design Document

phased implementation directly from a structure outline, skipping plan generation - you must use this skill when asked to implement an outline file or when asked to implement from a structure outline in .tasks/*

phased implementation of a structured plan you must use this skill when asked to implement a plan file in .tasks/*

iterate on design discussion based on user feedback - do not use this if you already used create-design-discussion

iterate on implementation based on user feedback

iterate on implementation plan based on user feedback

iterate on a PRD based on user feedback - do not use this if you already used create-prd

iterate on research-questions doc based on user feedback. don't use if you already used create-research-questions

iterate on research document based on user feedback. This skill requires a path to a document and feedback from the human - dont use if you already used create-research

iterate on structure outline based on user feedback. Don't use if you already used create-structure-outline

iterate on a TDD based on user feedback - do not use this if you already used create-tdd

Turn a rough intent — "implement Stripe webhooks here", "design module Y" — into a prompt you can hand to an implementing agent, or use to draft a plan first. Use BEFORE building anything non-trivial, when the request is one sentence and the real requirements, constraints, and external-system facts are still unstated. Do NOT use for localized, well-specified changes (a fix, a rename, a log line).

Use when browser automation is needed: open sites, click/fill forms, log in, scrape pages, take screenshots, or test web apps.

Use when working with Slack: read/search/send messages, inspect channels/threads/users, add reactions, manage canvases, or mark read.

Use when generating PR descriptions in CI or other noninteractive runs.

Use when the user explicitly asks to be grilled, stress-test a plan/idea/design, or sharpen vague thinking through relentless interview. Do not invoke unless explicitly asked by the user.

Use when working with Linear issues/projects: create, update, query, comment, attach files, or inspect Linear URLs/tickets.

Use when working with Notion via ntn: call APIs, query databases, create pages, upload files, or deploy/manage workers.

Use when creating, editing, reviewing, or testing coding-agent SKILL.md files, especially when reusable behavior should apply across repositories, tools, workflows, or agent sessions.