Prompt Library

Claude will only call Investra tools when it understands the request is real-estate-related and routable to one of our tools. Three patterns work reliably:

• Lead with "Using Investra AI…" — anchors the conversation in the Investra context and tells Claude to prefer our tools over generic web search.
• Name the tool explicitly — e.g. "use search_properties to…", "run estimate_costs on…", "call skip_trace for…". Removes ambiguity when multiple tools could apply.
• Specify output format — "give me a dashboard", "compare as a table", "rank by cap rate". Claude composes the tool output into the shape you ask for.

The fastest way to validate your connection is working. Each of these completes in under 10 seconds.

Using Investra AI, use search_properties to find 10 cash-flowing single-family houses in Pittsburgh, PA under $200k.

Using Investra AI, call analyze_market on Tampa, FL and summarize the rental vs. appreciation outlook.

Using Investra AI, run estimate_costs on 123 Main St, Columbus, OH with 20% down, 30-year loan, and self-managed rental. Return the A–F grade, cash flow, cap rate, cash-on-cash, and DSCR.

Using Investra AI, use get_sales_comps on 4124 Vallhorn Ln, Charlotte, NC 28213 to pull ARV low/base/high with confidence.

Using Investra AI, get_rent_estimate for a 3-bed 2-bath 1,400 sqft in Mesa, AZ — show worst/base/best rent scenarios.

The pattern you'll use most for portfolio planning and market comparison. The core prompt from the launch demo:

Using Investra AI, use search_properties to give me an entire dashboard of all properties and cash flow in Pittsburgh, PA and the other top 5 cash-flow markets. For each market, return:
  1. 10 best cash-flow listings with price, beds/baths, sqft, estimated rent, monthly cash flow, cap rate, and A–F grade
  2. Market median rent-to-price ratio
  3. One-sentence summary of why the market ranks

Layout as a side-by-side dashboard comparing all 6 markets.

Other dashboard shapes:

Using Investra AI, use search_properties with cashFlowOnly=true to build a 2-column comparison of Cleveland, OH vs. Birmingham, AL — 15 listings each, ranked by cap rate.

Using Investra AI, run search_properties across Indianapolis, Kansas City, Memphis, Jacksonville, and St. Louis — limit=10 per market — then build a ranking table of the top 5 deals overall by monthly cash flow.

Paid plans unlock search_off_market, get_off_market_details, and (Pro Plus) skip_trace.

Using Investra AI, use search_off_market to find 20 motivated-seller leads in Hillsborough County, FL (city=Tampa, state=FL). Prioritize absentee owners, tax delinquent, and pre-foreclosure.

Using Investra AI, use get_off_market_details on 4124 Vallhorn Ln, Charlotte, NC 28213 to pull motivation score, equity estimate, and distress flags.

Using Investra AI, call skip_trace on 4124 Vallhorn Ln, Charlotte, NC 28213 to get the owner's phone, email, and mailing address. (Pro Plus — uses 1 lead credit.)

Using Investra AI, build a 10-lead outreach pipeline for Memphis, TN:
  1. search_off_market for motivated sellers (limit=10)
  2. For the top 5 by motivation score, run get_off_market_details
  3. For the top 3 highest-equity leads, run skip_trace
  4. Return as a CSV-ready table: address, owner name, phone, email, motivation score, equity.

estimate_costs accepts sophisticated overrides — interest rate, down payment %, loan term, self-managed flag, whether landlord pays utilities — making it a full sensitivity engine.

Using Investra AI, run estimate_costs on 123 Main St, Tampa, FL 33601 three times:
  1. Base case: 20% down, 6.5% interest, 30-year, self-managed
  2. Conservative: 25% down, 7.0% interest, landlord pays utilities
  3. Aggressive: 15% down, 6.0% interest, tenant pays all utilities

Return a side-by-side comparison of monthly cash flow, cap rate, cash-on-cash, and DSCR for all three.

Using Investra AI, use estimate_costs with type=flip on 123 Main St, Tampa, FL 33601 — rehabScope=medium — then get_sales_comps to pull ARV. Calculate the expected flip profit after closing costs, holding costs, and the 70% rule.

Using Investra AI, run a BRRRR underwrite on 123 Main St, Tampa, FL 33601:
  1. get_sales_comps for ARV
  2. estimate_costs with type=brrrr, rehabScope=heavy
  3. get_rent_estimate for rent sensitivity
  4. Show expected cash-out at 75% LTV refi, monthly cash flow post-refi, and cash-on-cash return.

Using Investra AI, call get_zip_market_data on 15212 (Pittsburgh) to pull HUD Fair Market Rent, Census ACS demographics, vacancy rate, owner-occupied %, YoY appreciation, $/sqft, unemployment, and school/crime/flood scores. Then summarize whether this ZIP is a buy, hold, or avoid for single-family rentals.

Using Investra AI, use analyze_market on Jacksonville, FL and break down the pricing, rental, supply, demand, and economy sections. Call out any red flags an out-of-state investor should know about.

Using Investra AI, call get_market_data to pull current 30-year fixed rates, housing starts, and the unemployment rate. Based on these, tell me whether rates favor buying now or waiting 3 months.

Use find_local_partners to source vetted professionals in a new market before you buy.

Using Investra AI, I'm buying my first rental in Birmingham, AL. Use find_local_partners to pull the top 5 property managers, top 5 inspectors, and top 5 mortgage lenders in the market. Rank by rating and format as contact cards.

Using Investra AI, use find_local_partners to find 5 general contractors in Cleveland, OH — I need someone who can handle a $40k rehab on a 3-bed 1.5-bath.

Using Investra AI, I'm running a BRRRR in Kansas City, MO. Pull me:
  - 3 local lenders who do DSCR / portfolio loans
  - 3 property managers with 4.5+ ratings
  - 3 inspectors
Return as a ready-to-call list with phone and address.

Multi-tool workflows Claude can execute from a single prompt.

Full deal memo from one address

Using Investra AI, write a 1-page deal memo for 123 Main St, Tampa, FL 33601:
  1. get_property_details for the core facts
  2. get_sales_comps for ARV
  3. get_rent_estimate for rent sensitivity
  4. estimate_costs for cash flow, cap rate, cash-on-cash, DSCR
  5. get_zip_market_data for neighborhood risk
  6. Recommendation: Buy / Pass / Renegotiate — with reasoning.

Weekly deal hunt

Using Investra AI, every Monday I want you to:
  1. use search_properties with cashFlowOnly=true across Pittsburgh, Columbus, Indianapolis, and Memphis — limit=5 each
  2. Rank all 20 by cap rate
  3. For the top 5, run estimate_costs with 20% down and 7% rate
  4. Flag anything with DSCR > 1.25 and cash flow > $300/mo as "pursue"

Start now with this week's scan.

Portfolio stress-test

Using Investra AI, here are my 4 rentals: [addresses]. For each one:
  1. get_rent_estimate to see current market rent vs. what I'm charging
  2. estimate_costs at today's rates to see if refinancing makes sense
  3. get_sales_comps for current market value

Summarize: which properties are underrented, overvalued, or candidates for refi?

Claude answers from general knowledge instead of calling Investra.

• Re-prompt with the explicit tool name: "use search_properties to…", not just "find houses".
• Prefix with "Using Investra AI" to anchor the context.
• Check the MCP connection: /mcp in Claude Code, or the MCP panel on claude.ai — Investra should show 17 tools.

Tool returns an error or empty result.

• Verify your plan includes the tool (off-market and skip_trace need paid tiers — see pricing).
• Check your monthly quota at GET /api/usage or on your dashboard.
• For addresses, include the full street, city, state, ZIP — partial addresses often fail the lookup.

"Tool results could not be submitted — connection interrupted."

• The tool call exceeded Claude's MCP client timeout (~90s). Most common cause: asking for many markets or many results in one message.
• Fix: split the request into 2–3 messages, or reduce limit per market to 10 or lower.
• Refresh the conversation and retry once. If it happens repeatedly on a single-market single-tool call, check Troubleshooting.