MCP Prompts Overview
Phoenix exposes MCP prompts (workflows) that guide AI assistants through structured, repeatable research tasks. A prompt combines a templated message body with a small set of named arguments. Clients discover prompts via prompts/list and render them via prompts/get.
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
Start here: the getting-started prompt
The best way to begin is the getting-started prompt — a guided first run that takes you from
"just connected" to a real, rendered result. Pick it from your MCP client's prompt menu and it walks
you through five steps:
- Detect what's available. Phoenix looks at the tools in your current session — that set defines what it can run for you. Your API key is user-scoped: it never asks you to do org or admin setup.
- Ask two questions. Your role (Sales, Marketing, Customer Success, Exec / Strategy, or Other) and the company or product you represent. Nothing more — no open-ended questionnaire.
- Recommend one to three workflows. From the curated set below, Phoenix picks the best few for your role and the tools you have available.
- Run one live. Phoenix runs an Account Research Brief right away with realistic example inputs for your company, shows you the result, and renders the onboarding launchpad so you can see all your options at a glance.
- Offer one next step. A single, specific follow-up suggestion so you always know what to do next.
Curated workflow prompts
The getting-started prompt recommends from these ten curated GTM workflows. Each is a complete,
repeatable research task you can ask Phoenix to run on its own:
| Workflow | When to use it |
|---|---|
| Account Research Brief | A one-page brief — firmographics, tech stack, IT spend, and stakeholders — before your first call. |
| PVP / PQS Qualification | Score an account against your priority-vendor and product-fit framework to decide if it's worth pursuing. |
| Intent Targeting & Activation | Turn a buying-intent topic into a prospecting list of accounts actively in-market. |
| Vendor-Sprawl Consolidation Map | Map an account's overlapping tools to surface displacement and consolidation angles. |
| Pre-Call Brief | Focused prep on the account or contact you're about to talk to. |
| Competitive Analysis Brief | A side-by-side read on the competitive landscape and where you fit. |
| TAM Sizer (Tech Adjacency) | Size a market using technology-adjacency signals to find your real addressable base. |
| Competitive Battlecard | Objections, proof points, and traps for a named competitor. |
| ICP Refiner (Closed-Won Cohort) | Refine your ICP from the cohort that actually closed. |
| Market Analysis Brief | A sized, segmented, sourced read on a new segment's dynamics and where to play. |
Detailed, per-workflow documentation for each of these prompts is on the way. For now, the
getting-started prompt is the front door — it recommends the right ones for you and runs your first
one with you.
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
MCP clients render available prompts in a UI affordance (a slash menu or prompt picker). Invocation goes through the standard MCP prompts/list and prompts/get flow — no Phoenix-specific protocol.
Prompt Output Format
Prompts return a single user-role message containing the rendered template, with caller-supplied arguments substituted into {{varName}} placeholders. The rendered output is bounded; an oversized rendering (≥ 100 KB) returns an error rather than a giant message.