openclaw-workspace-report

Summary

This workspace documents agent roles, memory structure, tool usage, and a curated set of OpenClaw skills to support structured problem-solving, long-context operations, and automation tasks in a Python-centric environment. It is designed for an AI assistant operating in the "Kimi" style—prioritizing deep analysis, structured reporting, and efficient workflows, particularly for supporting the user 董兴屹.

Included Assets

• AGENTS.md
  Agent onboarding and session startup protocol, outlining roles, workflow, and reporting standards.
• MEMORY.md
  Defines the agent persona, user preferences, and search tool configuration.
• SOUL.md
  In-depth description of the assistant’s identity, answering style, core strengths, and operational philosophy.
• TOOLS.md
  Tooling reference for search, document handling, file operations, and integrations such as Obsidian.
• skills/database-designer/
  • README.md, SKILL.md — Documentation for a comprehensive database analysis and migration skill.
• skills/openclaw-github-assistant/
  • README.md, package.json — Documentation and metadata for GitHub repository management integration.
• skills/apple-reminders/
  • SKILL.md, metadata — Instructions and requirements for managing Apple Reminders via remindctl CLI.
• skills/coding/
  Documentation and templates for storing and applying user coding preferences, including:
  • SKILL.md, criteria.md, dimensions.md, memory-template.md

How to Use

1. Review Roles & Workflow

   • Start in AGENTS.md to understand expected behaviors, report structures, and session routines.
   • Reference SOUL.md for operational principles and correct assistant "personality".

2. Agent Session Handling

   • At the beginning of each session, follow the workflow described in AGENTS.md—read relevant memory files and user instructions.
   • Capture important context in daily memory logs and curate in MEMORY.md.

3. Apply Tools

   • Use the documented search and file utilities from TOOLS.md for automation tasks.
   • Only employ supported skills/tools as described (see notes on deprecated items).

4. Leverage Skills

   • Database design and analysis: See skills/database-designer/
   • Coding preferences and memory: See skills/coding/
   • GitHub repository management: See skills/openclaw-github-assistant/
   • Apple Reminders CLI integration: See skills/apple-reminders/

5. Update & Maintain

   • Log updates and lessons in the appropriate documentation or memory files per processes outlined (AGENTS.md and MEMORY.md).
   • Reference coding style preference criteria and templates when customizing to user feedback.

Notes

• The workspace enforces strong separation between user-provided explicit preferences and automated learning—especially for code style and behavior.
• Deprecated tools and workflows are clearly marked (e.g., certain search APIs are disabled).
• Designed around a single-user focus (董兴屹) and the "Kimi" assistant style: concise, structured, and context-heavy support.
• Integration with external tools (e.g., GitHub, Apple Reminders) requires proper environment setup and authentication as per each skill’s documentation.
• No external dependencies are required unless stated by individual skills (e.g., Node.js for GitHub skills, remindctl for Apple Reminders).
• This workspace is intended for agent-driven automation and documentation, not for building or extending agent code directly.