add_to_watchlist
ChatGPTAdd a property to the user's watchlist (their hand-picked pipeline, shown at /watchlist). Free; capped per plan.
Overview
Circlemark turns the county record into a working prospect list for commercial real estate. Ask for the loans about to move — by maturity window, borough or county, size, lender, property type or recorded distress — and get them property by property, with lender, borrower, amount and maturity taken from the recorded instrument. Every date carries its provenance: Circlemark says which maturities are stated on the document, which are a rate reset read from the recorded rider, and which are modelled, and it never presents an estimate as fact. Ask who lends on deals like this and it ranks the institutions that have closed comparable loans in a market, county, property type and size band over the last 36 months, each dated to its last closing — revealed appetite from recorded liens, never a quoted credit box. Then find the people. One reveal names the officers, owners and signers the public record ties to a borrowing entity — each cited to the source it came from, whether an assessor roll, a mortgage, an HPD registration or the signature page of the instrument — and adds verified work emails. Direct phone numbers are a separate, explicit request. Coverage today is New York City (all four ACRIS boroughs) and Texas — Dallas–Fort Worth, San Antonio and Austin — re-checked weekly. A recorded satisfaction or release marks a loan paid off and excludes it by default; matured-but-open, modified, foreclosed and distressed loans are flagged. Requires a Circlemark account; sign-in happens on circlemark.ai over OAuth. The assistant confirms with you before any contact reveal is recorded against your account.
Tools
find_lenders
ChatGPTThe institutions that have closed commercial mortgages matching a market, county, property type and loan-size band in the last 36 months, counted from recorded instruments (one per loan, originators only, three closings or more). Free (same daily fair-use meter as search_loans). This is revealed appetite, never a credit box: rank is by closings or recency, every row is dated to its lastDeal, and
stated / contacts are null because the record holds neither. Answers 'who lends on $5M industrial in Tarrant' with no subject property; for one property's comparables, search_loans first and pass its county, type and amount here.find_phone
ChatGPTRun the direct-phone lanes for one person on an already-unlocked property. Holds 5 credits; the charge settles only when a number lands (refunded on a miss). personKey comes from a prior reveal_contacts result. Some vendors answer by webhook — a 'pending' person means retry reveal_contacts in a few minutes to pick up the number.
phoneOutcome says what happened: 'none' = searched, no direct number; 'unavailable' = no source could search this person right now (nothing was looked up — say so, never 'no number exists'); 'error' = a lane failed, retry later.get_credit_balance
ChatGPTThe seat's reveal-credit allowance (this month's on a paid seat; the one-time budget on a free seat —
resets says which), usage, and the account's plan.get_lender
ChatGPTOne institution's commercial-mortgage record across every market Circlemark reads: closings by county, property type and quarter over the last 36 months, size band, all-time count, the recorded spellings folded onto it (a bank's old names stay on the clerk's index after a merger), its sibling desks per the reviewed crosswalk, and its last 20 closings. Free (same meter as a property view). Takes a groupKey from find_lenders or a name; an ambiguous name returns the candidates instead. Revealed behaviour, dated —
stated and contacts are null. The people at the lender are on the website profile page only, never through the connector.list_watchlist
ChatGPTThe property ids on the user's watchlist, newest pin first. Fetch details with search_loans or the property page.
remove_from_watchlist
ChatGPTRemove a property from the user's watchlist.
reveal_contacts
ChatGPTUnlock the people behind a property's leading loan: principals from the public record (each row cites its source) plus work emails from vendor lookups. Spends 1 reveal credit from the seat's allowance (monthly on a paid seat; a one-time budget on the free seat) — unless this property was already unlocked, in which case it is free and served from the earlier answer. Use propertyId from search_loans. Phone numbers are NOT included; use find_phone for one person.
search_loans
ChatGPTSearch commercial real-estate mortgages by maturity window, market, county, loan size, lender, and distress signals. Free to call (daily fair-use metering); returns one row per property × loan. statedMaturity is read from recorded documents; estimatedMaturity is a modelled estimate (maturityBasis 'assumed', or 'stated_term' when the document states a term but no date) and must be presented as an estimate, never as fact; nextResetDate (maturityBasis 'reset') is a stated rate-change date from the recorded rider, the event on 30-year paper. total counts to a depth of 1,000 ('1000+' beyond that).
Capabilities
Writes
Example Prompts
Click any prompt to copy it.
