Chatwin Business Diagnosis, Content Growth, and Multi-Role Chat Toolkit

What problem it addresses

When users face mixed problems—unclear business model, weak content, poor openings, AI-sounding drafts, scattered decisions, or messy local materials—jumping between separate tools creates context loss. chatwin acts as a routing entry point: it reads the current conversation, identifies the most actionable next step, and delegates to a specific sub-skill such as chatwin-diagnosis, chatwin-benchmark, or chatwin-content. Instead of showing a fixed menu, it navigates based on prior conclusions: 'knowing but not acting' routes to chatwin-action; vague concepts route to chatwin-deconstruct; weak openings route to chatwin-hook or chatwin-xhs-title.

How it works

The core loop is signal detection, single-step routing, sub-skill execution, and result-driven navigation.
- Pre-task routing: extract goals, materials, and constraints from phrases like 'pricing is unclear,' 'why did this go viral,' or 'is this script smooth,' then match a sub-skill.
- Post-task navigation: read the previous skill's conclusions, choose only the most relevant next direction, and continue immediately without asking the user to re-enter commands.
- Onboarding mode: on first use, explain what can be delivered, how it is processed, and how to start, then immediately complete a real first task.
- State continuity: support saving, restoring, and reporting diagnosis states through chatwin-save, chatwin-restore, and chatwin-report, keeping decisions and outputs locally recoverable.

It is useful for engineers and operators turning ambiguous business questions into content, titles, scripts, benchmarks, and decision records. The boundary is clear: chatwin routes and connects tasks; sub-skills do the actual diagnosis. Out-of-scope requests are rejected, and each step handles only the current best next action rather than forcing a long fixed chain.