Queen Creator Architecture Upgrade Wizard
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @user_5982371c/queen.
About this skill
Problem
queen targets architecture upgrades for OpenClaw Agents, focusing on three gaps: weak behavioral control, insufficient historical grounding, and limited long-term capability accumulation. It is not just prompt polishing; it splits harness control, memory structure, and evolution into executable upgrade layers.
How It Works
- It reads
SOUL.md,AGENTS.md, andMEMORY.mdto infer current Agent state, then appends a progress template. - It advances through STATE_H, STATE_M, and STATE_L: control focuses on avoiding errors and loss of control; memory focuses on retaining context with historical evidence; learning focuses on improving with use.
- After each layer, it updates the
MEMORY.mdprogress table and schedules a cron reminder for the next layer, supporting resume phrases like “continue queen upgrade”. - If base files are missing, it can suggest running
king.skillfirst, while still adapting independently from answers.
Boundaries
It suits engineers with an existing OpenClaw workspace who can absorb layered upgrades over days. It relies on cron, MEMORY.md, and reference files; without a basic Agent structure, behavior depends on follow-up answers. The optional iWiki sync branch should not be treated as a core requirement.
Use Cases
- Plan and run the Harness, memory, and learning upgrade layers for an existing OpenClaw Agent workspace.
- Advance the three queen layers on a 3-day or 7-day interval and create the next reminder after each layer.
- Resume the next queen layer from the last completed row in the MEMORY.md progress table.
- Sync the AutoDream report and upgrade progress to the iWiki knowledge base after completing the three layers.
Best For
- Engineers maintaining OpenClaw Agent workspaces who need to track control, memory, and learning upgrades step by step.
- Agent architects responsible for team structure who need a progress table and staged reminders for existing Agents.
- Engineers using MEMORY.md for project state who need to resume the next upgrade layer from historical records.
- Tencent engineers using iWiki who need to sync AutoDream reports and upgrade progress into the knowledge base.
Related Skills
Local workflow memory with matching and SOP updates.
An OpenClaw live streaming executor that initializes TRTC streaming, starts a real-time dashboard, generates viewer URLs, and continuously reports live events.
Breaks down physical supply chains for super-trends to identify second- and third-layer bottlenecks, runs valuation and reverse checks, and maintains trackable reports.
A token-saving compression mode for Chinese LLMs with lite, full, ultra, and classical tiers, preserving code and technical terms while handling edge cases.