In the research workflow of empirical asset pricing and machine learning / deep learning, literature search, data processing, model training, report writing, and presentation preparation are often independent tasks. DeepSeek Harness (DSH) connects these stages through a plugin mechanism. This section introduces Z-Asset/ZA_literature, which integrates literature review, search, synthesis, and citation chain tracking into a single plugin to support research workflows in asset pricing + ML/DL.

This is a literature review plugin responsible for search, synthesis, and citation chain tracking. It is part of the asset pricing + machine learning / deep learning research workflow and is maintained by Z-Asset.

Core Features

The plugin provides the following core capabilities:

  1. Literature review
  2. Data processing
  3. Model training
  4. Report writing
  5. PPT creation

Installation and Enablement

Install it using the DSH plugin command:

dsh plugin --profile <你的profile> add zresearch-literature

After installation, trigger it as needed in a DSH session. The model will automatically load the corresponding skill based on the task.

Typical Usage

  • Literature review: ask “help me do a literature review / find related papers”
  • Data processing: ask “help me find data / evaluate datasets”
  • Model training: ask “run a regression / train an ML model / cross-sectional return prediction”
  • Report writing: ask “write a paper / draft a chapter”
  • PPT creation: ask “create a presentation PPT / Beamer / job market talk”

Dependencies and Structure

The plugin depends on @deepseek-ai/cordis and @deepseek-ai/dsh-skill. Its skill structure includes skills/zresearch-literature/, with the main definition file SKILL.md and a references/ directory containing agents/ (worker/critic roles) and domain/ (domain calibration configuration).

Applicable Scenarios and Domains

The plugin targets the empirical asset pricing + ML/DL domain, covering factor models, cross-sectional return prediction, GMM/SDF, predictive regressions, machine learning, deep learning, and related areas.

Notes

The plugin runs with the permissions of the current DSH process. Before installation, please check the source code and license (MIT). It is designed specifically for asset pricing + ML/DL research scenarios and may not perform as expected in unrelated domains.

Directory and Source Code

Plugin directory: https://www.skillhub.cn/plugins/Z-Asset/ZA_literature
GitHub repository: https://github.com/Z-Asset/ZA_literature