Use when handling a customer's support message, retention care, VIP account management, goodwill grant, or live-event inquiry across a SaaS, AI/cloud or API service, casual game, heavy-IAP game, or consumer app: sign-in or recovery failures, missing…
Use when drafting, designing, comparing, or reviewing a launch/migration/rollout plan, system architecture, service boundary, data model, tech stack, or product spec; when stress-testing a commercial contract, reseller/partner agreement, pricing model,…
Use when implementing or shipping a feature, bug fix, parser, export format, endpoint, or authorized operational, repository, marketing, or commercial change. Carry the outcome through execution and verification at its real surface. Use self-challenge for a…
Use ONLY when explicitly asked to write, file, structure, or revise a work outcome item, issue ticket, outcome brief, or backlog entry (Stage 1 ATDD: defining goal, observable positive and negative acceptance test cases, constraints, and references without…
Use when creating, refactoring, or auditing a reusable Agent Skill (SKILL.md and its colocated references/*.md): write concrete moment-based trigger descriptions, package opinionated methodology (how to think, discuss, build, choose tech stacks, score risks,…
Use when drafting, rewriting, editing, translating, or localizing any text for clarity, natural tone, or persuasion: customer-facing announcements, release notes, blog posts, articles, marketing and landing-page copy, partner/studio proposal emails, UI…
Use when investigating a root-cause claim for a past failure; reviewing a PR, diff, proposal or contract for concrete defects; verifying a live status or metric before reporting it; reconciling conflicting readings; or checking that a resource or bug is…
Use when splitting an oversized module or test suite, deduplicating implementations, deleting dead code or unused dependencies, removing completed compatibility layers, or restructuring a codebase while preserving its observable behavior. Use…
Use when several independent views can change a decision: a multi-domain design, uncertain cause, architecture tradeoff or blocker needing new evidence. Assign distinct evidence angles and let the accountable owner decide. Use direct single-agent work for a…
Use when restoring a task, deployment or CI/build run that has actually stalled or is failing: a promised fix still fails, a call is refused, authority is missing, a runner hits a resource limit, or a queue/review stops progressing. Use prove-it and…