deep-research

To automate in-depth research and report generation, this Skill uses OpenAI and Google Gemini Deep Research APIs to process queries and generate comprehensive outputs. It manages long-running tasks asynchronously so you can start a query and poll for the completed report later.

Asynchronous execution for AI research tasks works through a subcommand architecture with start and poll commands. This prevents timeouts on long-running tasks by allowing you to initiate the research process and check back for results once the provider has finished generating the report.

Yes, you need API keys for both OpenAI and Google to run deep research tasks. The Skill requires these credentials to access the respective Deep Research APIs and execute the automated query processing and report generation workflows.

The system chooses between OpenAI and Google Gemini using smart provider selection based on query analysis and learned usage patterns. You can also manually select a strategy prioritizing either quality or cost to force a specific provider for your research task.

If an AI provider fails during the research process, the system uses robust error handling with retry logic and fallback mechanisms. It also provides detailed error logging to ensure reliable operation and continue your automated report generation workflow where possible.