Skip to main content

MCP Prompts Overview

Phoenix provides intelligent prompts that guide AI assistants through complex research workflows. Prompts combine multiple tool calls into structured, repeatable processes.

What are MCP Prompts?​

MCP prompts are pre-built workflows that:

  • Chain multiple tool calls together
  • Provide structured templates for research
  • Ensure consistent output formatting
  • Reduce complexity for common use cases

Available Prompts​

research_rule_generator​

Generates comprehensive research workflows for Account Executive (AE) teams.

Purpose: Create structured research rules that combine company data, technology insights, and intent signals into actionable sales intelligence.

Use cases:

  • Building prospect research workflows
  • Creating repeatable qualification processes
  • Standardizing sales intelligence gathering
  • Generating custom research templates

What it does:

  1. Analyzes your research requirements
  2. Identifies relevant Phoenix tools to use
  3. Creates a structured workflow
  4. Provides sample queries and expected outputs
  5. Generates documentation for your team

Example usage:

Use the research_rule_generator prompt to create a workflow for:
"Find enterprise SaaS companies using Salesforce with recent intent signals"

Output format:

  • Research rule name and description
  • Step-by-step tool sequence
  • Parameter recommendations
  • Expected data fields
  • Filtering and qualification criteria
  • Output formatting suggestions

Prompts vs Tools​

When to use Prompts​

  • Complex multi-step research workflows
  • Repeatable processes that combine multiple tools
  • Guidance on how to structure queries
  • Template generation

When to use Tools directly​

  • Simple single-step queries
  • Custom workflows not covered by prompts
  • Real-time data retrieval
  • Exploratory research

Using Prompts with MCP Clients​

Claude Desktop​

"Use the research_rule_generator prompt to help me build
a workflow for finding companies in healthcare using AWS"

Cline​

/phoenix research_rule_generator
Target: Healthcare companies with cloud migration signals

Other MCP Clients​

Invoke prompts using the standard MCP prompt protocol with the prompt name and required arguments.

Prompt Arguments​

research_rule_generator​

Required arguments:

  • objective: The research goal or question to answer
  • targetAudience: Who will use this research rule (e.g., "AE team", "SDRs")

Optional arguments:

  • includeIntentSignals: Whether to include intent data (default: true)
  • spendAnalysis: Whether to include spending data (default: false)
  • technologyFilters: Specific technologies to focus on

Example:

{
  "objective": "Find mid-market companies showing cloud migration intent",
  "targetAudience": "Enterprise AE team",
  "includeIntentSignals": true,
  "spendAnalysis": true,
  "technologyFilters": ["AWS", "Azure", "GCP"]
}

Best Practices​

1. Be Specific​

Provide clear, specific objectives to get the most useful research rules.

❌ Bad: "Find companies" ✅ Good: "Find Series B SaaS companies with 50-200 employees using Salesforce and showing recent CRM buying intent"

2. Define Your Audience​

Specify who will use the research rule to get appropriately detailed outputs.

Examples:

  • "SDR team" → Simpler, action-focused workflows
  • "Enterprise AE team" → More detailed, multi-touch research
  • "Marketing team" → Campaign-focused, segmentation-heavy

3. Iterate on Results​

Start with a basic prompt, review the generated workflow, then refine your requirements.

4. Combine with Tools​

Use prompts to generate workflows, then execute them with individual tools for actual data retrieval.

Prompt Output Format​

Prompts return structured markdown with:

  1. Research Rule Overview

    • Name and description
    • Target use case
    • Expected outcomes

  2. Tool Sequence

    • Ordered list of tools to call
    • Parameters for each tool
    • Data flow between tools

  3. Filtering Criteria

    • How to qualify results
    • Threshold values
    • Decision logic

  4. Output Template

    • Expected data fields
    • Formatting recommendations
    • Example outputs

  5. Usage Instructions

    • How to execute the workflow
    • Tips for your specific use case
    • Customization options

Example Workflows​

Enterprise Sales Research​

Prompt: research_rule_generator
Objective: "Identify enterprise companies (>1000 employees) in financial
services using Salesforce with recent intent signals for data analytics tools"
Audience: "Enterprise AE team"

Generated workflow:

  1. Use company_search to find financial services companies >1000 employees
  2. Filter for those using Salesforce with company_technographic
  3. Check intent signals for data analytics topics with company_intent
  4. Get detailed company profiles with company_firmographic
  5. Prioritize by intent score and company size

Technology Migration Research​

Prompt: research_rule_generator
Objective: "Find companies migrating from on-premise to cloud infrastructure"
Audience: "Cloud solutions team"
Include: Intent signals and spend analysis

Generated workflow:

  1. Search for companies with legacy infrastructure
  2. Check cloud spending trends
  3. Identify migration intent signals
  4. Analyze current technology stack
  5. Score migration readiness

Coming Soon​

Additional prompts in development:

  • competitive_analysis - Generate competitive intelligence workflows
  • market_sizing - Create market size estimation processes
  • account_expansion - Build upsell/cross-sell research rules
  • territory_planning - Design territory research workflows

Next Steps​