SuperAPI AI Directory
Partner中文AI newsFavorites
Skillsdeepresearch-v2

deepresearch-v2

Writing & research

View repository

Updated 10-16 03:35

About this skill

Please deep think / ultrathink as this is a complex task.

deepresearch-v2

Craft a structured research prompt to leverage a deep research agent for solving coding challenges. Domain-aware: surfaces boundary concerns and cross-domain impacts when relevant.

Pre-flight: Load Domain Context

Before crafting the prompt, check for domain context:

  • If docs/domains/registry.md exists → scan registered domains
  • If docs/domains/domain-map.md exists → note domain relationships and contract edges
  • For domains related to the problem → read docs/domains/<slug>/domain.md (concepts, contracts, composition)

If no domain system is present, skip this step and proceed normally.

Crafting the Research Prompt

Please create a structured prompt with the following sections:

1. Clear Problem Definition

  • Precisely define the issue we are facing.
  • Include relevant error messages, stack traces, or problematic behaviors.
  • Which domain(s) does this problem affect? Does it cross domain boundaries?
  • If multi-domain: which domain owns the fix vs. which domains are impacted?

2. Contextual Information

  • Specify the technology stack involved (languages, frameworks, libraries).
  • Include exact versions of key dependencies and any recent updates.
  • Mention any recent codebase changes potentially related to the problem.
  • Note domain-specific dependencies: contracts consumed, contracts exposed, shared infrastructure.
  • Flag domain separation concerns — is the problem caused by tight coupling across boundaries?

3. Key Research Questions

  • Identify clear, actionable research questions to guide the deep research agent.
  • Ensure these questions target resolving our current challenge directly.
  • If domains are involved: Does the solution respect existing domain boundaries? Should a contract be introduced or modified?

4. Recommended Tools and Resources

  • List specific tools, libraries, or methodologies we should investigate.
  • Include any helpful links or references.

5. Practical Examples

  • Request practical code examples demonstrating the recommended solutions.

6. Pitfalls and Mitigation

  • Ask for common mistakes related to the identified issue and best practices to avoid them.

7. Integration Considerations

  • Include any special considerations or potential impacts on our existing workflows and CI/CD pipelines.
  • Domain boundary impacts: Will the solution require changes to domain contracts? New cross-domain dependencies?
  • Cross-domain communication patterns: Does the solution need new events, shared types, or adapter layers between domains?
  • If new infrastructure is needed, does it belong in a _platform domain or the feature domain?

Please ensure the prompt is detailed, clear, and actionable, facilitating precise and valuable insights from our deep research agent.