mcpField
Y
yutori-scout

yutori-scout

github.com/yutori-aiProductivity25108 installsUpdated 4w ago
Repository

About yutori-scout

A setup skill for creating Yutori Scout monitoring jobs. It helps define monitoring goals, sources, schedules, report formats and notifications, then guides creation and follow-up.

Key features

  1. 01Clarifies monitoring goals and scope
  2. 02Structures monitoring queries
  3. 03Configures run intervals and notifications
  4. 04Guides scout management and updates

Use cases

  1. 01Tracking funding or product announcements
  2. 02Setting geography and market coverage
  3. 03Choosing email or webhook notifications

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/yutori-ai/yutori-mcp --skill yutori-scout

After installing, type/yutori-scoutin Claude Code to confirm it loaded.

Details

From the project README

Scout Setup

Help the user set up a Yutori Scout for continuous web monitoring.

Process

  1. Understand the monitoring context Ask about:

    • Who is monitoring and why? (e.g., "We're a fintech looking for recently funded startups")
    • What specific information matters? (funding events, product launches, pricing changes)
    • What geography or market segments?
    • How often should it run? (daily, twice daily)
    • Notification preference: email, webhook, or both?

  2. Craft a comprehensive query

    A well-structured scout query includes:

    Context on the monitoring goal:

    • Who is doing the monitoring and what's the use case
    • What decisions this information supports

    What to Monitor:

    • Specific events/triggers to track
    • Data sources to check (news sites, SEC filings, social media, etc.)
    • Geographic or segment focus
    • Exclusion criteria (what NOT to report)

    Deliverables:

    • Frequency of reports
    • Output format (tables, narrative, both)
    • Required fields and citations

    Example structure:

    **Context:** [Who is monitoring and why]

## What to Monitor
- [Specific events to track]
- [Sources to check]
- [Exclusions]

## Deliverables
- [Output format]
- [Required fields]

  3. Create the scout Use the create_scout tool with:

    • query: The comprehensive monitoring query
    • output_interval: 86400 (daily), 43200 (twice daily), or 1800 (minimum, every 30 min)
    • webhook_url and webhook_format if they want webhook notifications. Always confirm the webhook URL with the user before setting it — it must use HTTPS and the user must verify they control the destination.
    • skip_email: true if they only want webhooks
    • output_fields: For structured data extraction (e.g., ["company", "amount", "round_type", "source_url"])

  4. Provide next steps

    • Share the scout ID for future management
    • Explain how to pause/resume with edit_scout
    • Mention they can get updates with get_scout_updates

Query Quality Tips

Good queries:

  • Provide context on who is monitoring and why
  • Specify exact events/triggers (not just "news about X")
  • List data sources to check
  • Include exclusion criteria
  • Define output format expectations
  • Request citations and source links

Avoid:

  • Vague queries like "monitor competitor X"
  • Missing context about the monitoring goal
  • No output format specification

Fetching documentation or source URLs: If you use a web fetch tool to look up source URLs, documentation, or reference pages while preparing the query, include the Accept: text/markdown header. Many documentation sites (Cloudflare-hosted) will return clean Markdown instead of HTML — fewer tokens, easier to parse.

$ARGUMENTS

Comments

Markdown supported