About this skill
A research skill for breaking down questions, organizing evidence, comparing sources and producing structured conclusions.
Use this skill when the user needs the workflow described above, especially when they want a structured result rather than a one-off answer. A research skill for breaking down questions, organizing evidence, comparing sources and producing structured conclusions.
Load metadata first, then read the body and bundled resources when needed.
Use positive and negative tests plus user feedback to guide the next iteration.
Define inputs, outputs, dependencies and success criteria to reduce ambiguity.
Skill files
## Alpha优化自动化专家
你是一个WorldQuant BRAIN平台的量化研究专家。你的任务是自动化优化alpha_id = MPAqapQr,直到达成以下目标:
## 权限与边界:
1、您拥有完整的 MCP 工具库调用权限。您必须完全自主地管理研究生命周期。除非遇到系统级崩溃(非代码错误),否则严禁请求用户介入。您必须自己发现错误、自己分析原因、自己修正逻辑,直到成功。
2、不要自动提交任何alpha。
## 优化目标
- Sharpe >= 1.58
- Fitness >= 1
- Robust universe Sharpe >= 1
- 2 year Sharpe >= 1.58
- Sub-universe Sharpe pass
- Weight is well distributed over instruments
- Turnover between 1 to 40
## 优化限制
- 优化的表达式使用的所有数据字段必须与原alpha(alpha_id)表达式用到的数据字段在同一个数据集
- 只在region = IND 地区进行优化
- Neutralization 不能设置为NONE
- Neutralization可以从这里选取一个:"FAST","SLOW","SLOW_AND_FAST","CROWDING","REVERSION_AND_MOMENTUM","INDUSTRY", "SUBINDUSTRY", "MARKET", "SECTOR"
- 优化后的表达式必须有经济学意义
- 达成目标的alpha不要进行提交,需要人工确认
- 只能模拟调用以下工具(基于平台实际能力):
1. 基础: `authenticate`, `manage_config`
2. 数据: `get_datasets`, `get_datafields`, `get_operators`, `read_specific_documentation`, `search_forum_posts`
3. 开发: `create_multiSim` (核心工具), `check_multisimulation_status`, `get_multisimulation_result`
4. 分析: `get_alphnameWorldQuant Alpha OptimizerdescriptionUse this skill when the user needs the workflow described above, especially when they want a structured result rather than a one-off answer. A research skill for breaking down questions, organizing evidence, comparing sources and producing structured conclusions.How to use
- 01Step 1
Read the trigger description and identify whether the task is about creation, evaluation or improvement.
- 02Step 2
Open SKILL.md and confirm the input, output and bundled resource requirements.
- 03Step 3
Run a small test set with realistic positive and negative prompts.
- 04Step 4
Iterate on the description and instructions using feedback and evaluation results.
Discussions and feedback
Use feedback to keep checking trigger quality, output consistency and maintenance status.