Introduction

DeepSeek Harness (DSH) uses a plugin-based architecture, aiming to break down complex systems into composable modules. For individual users who want to transition roles, fill capability gaps, or plan a career path, three specific problems are common: not knowing what to do today (goals are too large), learning a lot but producing no tangible results (lack of evidence), and gradually drifting away from the direction (lack of review). The dsh-growth-workbench plugin connects these three areas into a closed-loop process: what to do today → leave evidence → periodic assessment → adjust based on results.

What It Is

This is an AI personal growth workbench for general users. It turns “I want to transition” into “one thing I can finish today.” The maintainer is mingdui, and it follows the MIT license.

Core Features

The plugin provides the following capabilities:

  • Profile: Establishes the user’s basic conditions and transferable skills through Q&A.
  • Capability Model: Analyzes capability gaps and generates a reinforcement order.
  • 90-Day Plan: Breaks long-term goals down into daily tasks.
  • Today Execution: Focuses on a single task each day and records completion.
  • Evidence: Records task completion status and supports images or files.
  • Assessment & Review: Includes milestone quizzes and stage exams.
  • AI Scoring: Scores results based on assessments.

Installation and Activation

The environment requirements are: Node.js ≥ 22.19, DSH CLI available in PATH, and a profile that supports bundle plugins (for example, web).

Install it using the official command:

dsh plugin --profile web add github:mingdui/dsh-growth-workbench

After installation, the DSH process must be restarted for the changes to take effect. You can confirm the installation with the following command:

dsh --profile web --dump-config | grep growth-workbench
# Expected output includes: - id: dsh-growth-workbench

Usage Flow

After restarting DSH, find “Growth Workbench” in the left menu.

  1. Profile: Choose a direction, answer identity-related questions (such as student/employed/freelance), confirm transferable skills and basic capabilities, and complete a self-assessment.
  2. Plan: Generate a 90-day plan based on the self-assessment results.
  3. Today: Open the “Today” tab each day, complete only one planned task, and write one sentence of evidence after completion (with an image or attachment if needed).
  4. Assessment: An assessment is triggered when the deadline arrives (milestone quiz or stage exam). Open the exam paper in the “Assessment” tab, and after submission, the AI will score it based on the conversation.

Applicable Scenarios and Notes

This plugin is suitable for users who need systematic planning for personal growth. When using it, pay attention to the following boundaries:

  • Capability Model: This is not a strict industry standard and is for reference only.
  • AI Scoring: There is no automated acceptance process, so the reasonableness of scores must be confirmed manually.
  • Reminders: There is currently no timed reminder feature, so assessment milestones must be tracked manually.
  • Security: The endpoints do not have independent authentication. The DSH web server binds to 127.0.0.1 by default. If you need to expose it externally, you must add an authentication layer yourself.

Summary

dsh-growth-workbench transforms vague personal growth goals into an actionable 90-day plan and daily tasks. Through the “profile-plan-execution-evidence-assessment” process, it helps users complete the entire workflow from goal setting to implementation review within the DSH environment.