[WIP] Start implementation in ndb-core repository

[Copilot]

• Create AGENTS.md file for the replication-backend repository with AI agent guidelines adapted from ndb-core, covering:
  • Project context (NestJS, CouchDB proxy, permissions)
  • TypeScript/NestJS development patterns
  • Testing guidelines (Jest unit tests, e2e tests)
  • Sentry.io integration guidance
  • GitHub MCP usage
  • Common commands
  • Agent mode workflow instructions

Original prompt

Start implementation in the ndb-core repository

The user has attached the following file paths as relevant context:

• AGENTS.md

[Chronological Review: The conversation began with a request to review the Agents.md file based on the latest models and best practices. The user outlined specific needs for agents, including fleshing out user requirements, troubleshooting, planning technical implementations, and generating tests. Subsequent messages involved discussions about the necessity of certain instructions, the inclusion of specific tools like Sentry.io and GitHub MCP, and whether to link relevant instruction files. The user then requested to start implementation, which was followed by a command to cancel and delegate to a cloud agent. Finally, the user requested to create a new branch from the master branch for the changes.]

[Intent Mapping:

1. Review Agents.md based on latest models and best practices.
2. Discuss the value of adding planning and architecture instructions.
3. Clarify Angular versioning and the necessity of Claude-specific workflow instructions.
4. Add Sentry.io MCP server and GitHub MCP for issue interaction.
5. Decide on the coverage of unit and e2e test patterns in Agents.md.
6. Link relevant instruction files from Agents.md sections.
7. Start implementation.
8. Cancel the implementation and delegate to a cloud agent.
9. Create a new branch from the master branch for changes.]

[Technical Inventory:

• Tools: Github Copilot, coderabbit.ai, Claude Sonnet, Claude Code.
• Technologies: Angular (latest version), Sentry.io, GitHub MCP.
• Concepts: Clean architecture, user requirement fleshing, troubleshooting, technical implementation planning, code refactoring, e2e testing.]

[Code Archaeology: No specific files or code changes were discussed in detail, but the conversation revolved around the structure and content of Agents.md and README.md.]

[Progress Assessment: The user has outlined the requirements for agents and discussed various aspects of the documentation. The implementation phase was initiated but then canceled in favor of delegating to a cloud agent. A new branch creation was requested.]

[Context Validation: All critical information for continuation is captured, including the need for documentation updates and the creation of a new branch for changes.]

[Recent Commands Analysis:

• Last Agent Commands: "Start implementation" and "Cancel: 'Delegate to cloud agent'" followed by "create a new branch from master branch for these changes."
• Tool Results Summary: No specific tool results were returned as the commands were related to documentation and branch management.
• Pre-Summary State: The agent was actively working on starting the implementation when the token budget was exceeded.
• Operation Context: The commands executed were part of the user's goal to implement changes and manage the project structure effectively.]

1. Conversation Overview: - Primary Objectives: Review and improve Agents.md based on latest models and best practices, clarify documentation needs, and manage project changes effectively. - Session Context: The conversation flowed from reviewing documentation to discussing specific technical needs, followed by implementation commands and branch management. - User Intent Evolution: The user shifted from wanting to start implementation to canceling it and delegating to a cloud agent, indicating a need for a more efficient approach.

2. Technical Foundation:

• Github Copilot: Used in VS Code IDE for code assistance.
• coderabbit.ai: Utilized for code reviews.
• Claude Sonnet/Opus models: Referenced for AI assistance.
• Sentry.io: Suggested for troubleshooting and monitoring.
• GitHub MCP: Mentioned for interaction with issues and pull requests.

3. Codebase Status:

• Agents.md:
  • Purpose: To provide guidelines for agent usage and best practices.
  • Current State: Under review for updates based on user requirements.
  • Key Code Segments: Discussions on user requirements, troubleshooting, and planning.
  • Dependencies: Relates to README.md for developer instructions.

4. Problem Resolution:

• Issues Encountered: Clarification on the necessity of certain instructions and documentation structure.
• Solutions Implemented: Suggestions to add Sentry.io and GitHub MCP, and discussions on linking relevant instruction files.
• Debugging Context: No specific debugging issues were mentioned, but troubleshooting needs were highlighted.
• Lessons Learned: Importance of clear documentation and the need for specialized agent instructions.

5. Progress Tracking:

• Completed Tasks: Initial review of Agents.md and discussions on documentation needs.
• Partially Complete Work: Implementation was started but then canceled; branch creation is pending.
• Validated Outcomes: No specific features confirmed working through testing yet.

6. Active Work State:

• Current Focus: The user was focused on starting the implementation of changes.
• Rece...

Created from VS Code.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.