core-principles

Use when planning work that needs multiple AI agents — designing agent roles, sizing the swarm to the task, and running agents in parallel or stages.

Use when working CTF challenges, sanctioned security exercises, hack-the-box style labs, or authorized penetration-testing tasks — covers scope discipline…

Use when reviewing an agent's development or complex code work — adversarial, question-driven review with authority to demand partial or full rework.

Use when a work session or major task completes — evaluate performance, extract reusable knowledge, and persist it to memory stores and project folders.

Use when writing a skill that enforces a rule agents skip under pressure.

Use when responding to feedback from a code review, whether from a subagent or human partner.

Use when working on confidential, NDA-covered, or otherwise non-public R&D projects, data, or client engagements.

Use when you have multiple independent tasks that can be executed concurrently by subagents.

Use before any creative work — creating features, building components, adding functionality, or modifying behavior.

Use when the user asks about attribution audit, attribution model or windows, AdAttributionKit, AAK, SKAN, view-through attribution, GA4 attribution, Consent…

Use when the user asks about budget allocation, bidding strategy, ad spend split, ROAS target, media budget, scaling campaigns, or which campaigns to kill.

Use when the user asks for competitor ads, ad spy, competitive analysis, competitor PPC, ad intelligence, Meta Ad Library, Facebook Ad Library, Google Ads…