cuecue-deep-research_skill

Overview

This skill conducts deep financial research using CueCue's AI-powered multi-agent system. It orchestrates agents to decompose a question, gather data from web and financial sources, synthesize findings, and produce a shareable markdown report. The process streams progress and can save a timestamped report file for later review.

How this skill works

The skill starts a background research job that coordinates supervisor, researcher, and reporter agents. It streams task-level progress (task titles, report generation status) to the console and produces a final markdown report and a report URL. For long runs, use the built-in progress URL and automated monitoring to receive periodic updates until completion.

When to use it

• Deep-dive company or sector financial analysis requiring multi-source research
• Competitive landscape or supply-chain risk studies that need synthesis and citations
• Preparing investor memos, executive summaries, or long-form research reports
• When you want a reproducible, timestamped markdown report saved to disk
• Cases where you need automated, periodic progress updates for long-running queries

Best practices

• Provide a clear, specific research question to reduce ambiguous results
• Use the recommended output path format ~/clawd/cuecue-reports/YYYY-MM-DD-HH-MM-name.md
• Configure CUECUE_API_KEY in your environment or platform config to avoid passing keys each run
• Monitor the command output every 5 minutes or enable automated cron monitoring for updates
• Use either a research template (template-id) or mimic-url—not both—to control report style

Example use cases

• Compare BYD vs Tesla market positioning and profit drivers; save a timestamped report
• Analyze Q3 revenue and margin trends for a public company with sourced citations
• Assess metal commodity price drivers and build an investment-ready summary
• Track progress of a long-running research job and receive periodic chat updates
• Mimic an organization’s writing style for consistent external reports