Generate parcel reports without turning diligence into surprise data costs.

1. 1Choose the market you want to evaluate, then start with a county, ZIP, city, or saved buy box.
2. 2Search a ZIP code or city to center the map. The location defines the market, not an automatic promise to pull every parcel.
3. 3Attach a marketplace or broker listing URL plus asking price when you want AcreSignal to compare the ask against parcel facts and underwriting.
4. 4Start with a free preview to confirm the parcel and see the public teaser without paid provider calls.
5. 5Run the first full report free when the parcel deserves deeper diligence; buy one-credit or two-credit packs for additional parcels.
6. 6Generate the full report, review the unlocked personalized synthesis, scores, assumptions, underwriting, and action plan, then download a buyer-ready PDF or send it straight to ChatGPT, with a QR/link for buyers, partners, or clients to ask the report questions.
7. 7Use the Compare Deals workspace to rank two or more unlocked reports by thesis fit, risk, underwriting, source coverage, and next action.
8. 8Add important reports to the watchlist when you want AcreSignal to monitor price, comps, hazards, imagery, owner, tax, or score changes.
9. 9Create area watches for counties, ZIPs, states, or custom buy boxes when you want stored opportunities ranked against a market thesis.

The free preview is intentionally narrow: it confirms the parcel context, shows public teaser signals, and avoids paid provider calls or full report persistence for anonymous viewers.

The first full report can run without a paid credit. Additional full reports run after a report credit is available, use the deeper provider stack, store reusable ledger observations, and produce a shareable report, a buyer-ready PDF, a one-click ChatGPT handoff, and report Q&A link.

AcreSignal accepts user-supplied listing URLs from marketplaces and broker pages as context for a parcel report. The app normalizes the URL, strips query strings and hashes, identifies the marketplace, and calculates price per acre when asking price and acreage are available.

The listing-intake layer does not scrape marketplace pages. `/api/openland/listing-handoff` is the safe bridge for Chrome-extension style workflows today: it returns a prefilled report-start URL, callable MCP preview arguments when parcel identity is present, and a paid-report queue template for assistant clients, CRMs, or browser sidecars.

The AcreSignal Sidecar release-candidate package is available in the app repo and documented at /sidecar. It reads only the current page URL, relies on user-entered price, acreage, APN, coordinates, and strategy, then opens the same report-start flow or copies an MCP payload.

Paid reports turn diligence into a disposition packet: target buyer, lead angle, listing narrative, proof metrics, buyer objections, visual asset plan, exit channels, and follow-up prompts.

This is the practical bridge from analysis to sale. The same packet can later receive broker shortlist data, listing-photo guidance, Google Aerial View or 3D presentation assets, and assistant-generated outreach without changing the report structure.

The visual plan is also available directly at /api/openland/reports/[id]/visual-assets and through MCP tool get_visual_asset_plan. Report owners can create private, rights-attested signed upload contracts with POST /api/openland/reports/[id]/visual-assets or MCP create_visual_asset_upload, then approve, reject, or archive uploaded assets through /review or MCP review_visual_asset. Approved image assets can be analyzed through /analyze or MCP analyze_visual_asset to produce buyer-packet evidence, captions, caveats, and edit briefs. An analyzed asset can queue, process, and review a generated/edited output through /generate, /generate/process, /generate/review, or the MCP tools queue_visual_asset_generation, process_visual_asset_generation, and review_generated_visual_asset. Generated media remains private until owner-approved with disclosure before artifact inclusion, and raw media stays out of anonymous public previews.

Report owners now get the same generated-output workflow inside the report page: private source previews, analysis state, generation state, owner-only generated previews, and approve/reject controls before buyer-facing PDF, ChatGPT/Markdown, or packet artifacts can use generated media.

The investor dashboard now turns stored opportunities into a campaign plan before export spend: ready records, review records, suppressions, scrub reasons, negative-keyword checks, owner-export columns, standard skip-trace estimate, premium skip-trace estimate, and DNC-screen estimate.

This does not buy owner records or skip traces by itself. It is the decision layer that keeps list spend tied to strategy fit, underwriting, scrub status, source coverage, and the investor's saved reject rules. Assistant clients can call the same workflow through the MCP tool build_campaign_export_plan.

The dashboard can export the stored campaign plan as CSV or Markdown. The MCP tool can also return a CSV or Markdown artifact by passing artifactFormat. These exports use stored opportunity data and intentionally omit owner/contact values until a licensed owner-export provider is activated.

The next controlled step is POST /api/openland/owner-export-jobs or MCP queue_owner_export_job. It creates an idempotent provider-gated owner-export request, stores a redacted workpaper, estimates owner-record spend, and keeps status at provider_required until a licensed provider handler is connected.

AcreSignal sends production lifecycle email through the verified Resend sender domain: report-ready delivery, final failed-report notice after retries are exhausted, automatic credit-return notice, report-credit purchase receipt, and direct report-unlock receipt.

Email sends are tagged and idempotent. Worker retries and Stripe webhook retries use stable keys so the same report job or checkout session does not intentionally send duplicate customer emails.

Reports include an AcreSignal investor synthesis with a deterministic fallback. It explains why the parcel matters for the selected strategy, what may be mispriced, what the underwriting says, which recurring-revenue paths deserve diligence, what risk tradeoffs remain, and which questions to ask the parcel next.

The synthesis can use the investor profile for fit scoring while keeping private buy-box values such as exact budget, target markets, offer discipline, reject rules, and risk tolerance out of shareable report text.

Unlocked reports and stored opportunities can be re-underwritten without new data charges. The authenticated API at /api/openland/deal-underwriting accepts a report or opportunity id plus an optional current asking price, then returns max offer, ask posture, spread, investor fit, risk, recurring-revenue paths, caveats, and next action from the ledger.

Assistant clients can call the same decision frame through MCP tool underwrite_deal, which keeps owner fields scoped to owner:read and lets investors test a live listing ask inside Claude, Codex, Cursor, or another bearer-token MCP workspace. ChatGPT remains an OAuth connector target.

The public endpoint /api/openland/market-score gives browser sidecars and assistant clients a low-cost market teaser from user-supplied listing context, state/county/ZIP, acreage, asking price, and strategy. It uses stored aggregate AcreSignal data only, adds no extra data cost, avoids storing anonymous score payloads, and suppresses the score when a market has too few stored samples.

AcreSignal tracks the same investor jobs land platforms market around: market selection, price-per-acre context, owner exports, scrubs, due-diligence overlays, skip tracing, comping, offer math, realtor packets, listing visuals, campaign workflow, and browser-side listing analysis.

The AcreSignal direction is more agentic: every report, listing, watched parcel, area watch, artifact, and accepted investor preference compounds into the ledger. That lets future market briefs reuse stored intelligence before spending on duplicate provider calls.

Premium source lanes include access-dependent Google Places Aggregate for amenity and buyer-demand density, Maps Grounding Lite for assistant-native location context, Population Dynamics for sibling-market modeling, Street View Insights for access and infrastructure evidence, Aerial/Satellite Insights for imagery change detection, Aerial View and 3D Tiles for buyer presentations, and image and map analysis for licensed imagery and report assets. These sources must follow provider attribution, storage, product-stage, and data-use limits.

The public source-coverage contract at /api/openland/data-coverage classifies where each data source appears across AcreSignal: mapped to implemented surfaces, mixed, provider-required, planned, or unmapped. It is not a runtime provider-health check. Assistant clients can call list_data_coverage through MCP before making claims about owner exports, comps, contact enrichment, Google geospatial AI, or premium imagery.

Licensed market coverage now has a separate internal lane for ZIP/county market geographies, active/pending/sold land comp snapshots, and derived market metric snapshots. Raw licensed rows are service-role only; public reports, sidecars, and MCP clients should expose aggregate summaries after sample suppression and provider-rights checks.

Customer-facing pricing uses report credits, not raw API calls or monthly subscriptions. The first full report is free for distribution and onboarding. After that, one report credit generates one full parcel intelligence report. The current launch offers are $79 for one extra report and $150 for two reports.

Internally, cost is driven by parcel data, map data, ownership enrichment, comp pulls, terrain and hazard layers, AI synthesis, artifact generation, and background processing.

The Compare Deals workspace evaluates stored reports side by side for a selected strategy. It does not trigger new paid provider calls; it reuses report scores, opportunity synthesis, underwriting, source coverage, and redacted listing context.

Use it after generating multiple reports to answer the practical investor question: which parcel wins for this thesis, what max-offer discipline should hold, and what downside checks still matter?

Paid report generation stores source-call timestamps and expiration windows in the intelligence ledger. Shared reports and MCP report briefs summarize which sources are fresh, expiring, stale, missing, or untracked without making new provider or AI calls.

This helps investors separate a durable parcel thesis from evidence that should be refreshed before offer math, owner outreach, buyer packets, grant claims, water planning, or solar underwriting. Saved area-watch lists, assistant responses, and full market briefs now also consume aggregate ZIP/county market metric snapshots when they exist, without exposing raw licensed comp rows.

The area-watch refresh route is guarded by an atomic Supabase refresh-budget ledger: per-run caps, daily watch caps, provider-budget caps, and explicit quota-skip observability. The current mode still adds no new data charges. Each successful refresh stores source-freshness and aggregate market-metric snapshots by saved geography so stale market evidence is visible before an investor acts.

AcreSignal can be connected to assistant clients through the MCP endpoint at https://acresignal.land/api/mcp. Create a scoped token from Settings, then add it as a Bearer token in the assistant or agent client.

Keep tokens narrow by default. Use report read access for summaries, queue-report access only when the assistant should run full reports, and owner-field access only for workflows that need sensitive seller or mailing data. Tokens are shown once, always expire, and sensitive report-writing or owner-field tokens are capped to short lifetimes; revoke any token that is exposed.

The public connection kit at /api/openland/mcp-client-config publishes copy-ready setup snippets for Claude remote MCP clients, Codex, Cursor, internal agent stacks, and generic HTTP MCP clients. It also marks ChatGPT as an OAuth-required connector target so production setup is not confused with bearer-token clients.

MCP auth discovery is available at /.well-known/oauth-protected-resource/api/mcp. It advertises scoped bearer-token support plus the OAuth authorization server, and keeps ChatGPT marked as customer-pending until connector packaging and approval are complete.

AcreSignal now exposes the same-origin OAuth authorization-server metadata at /.well-known/oauth-authorization-server, with browser authorization at /api/mcp/oauth/authorize and token exchange at /api/mcp/oauth/token. OAuth connector v1 is read-only, requires PKCE, stores authorization codes and access tokens only as hashes, and grants preview:read plus reports:read. Report-writing and owner-sensitive assistant access still require manually created private MCP tokens until the connector consent UX is expanded.

The same connection kit now publishes a ChatGPT connector launch packet with the intended connector name, connector URL, readiness gates, first-run prompts, read-only starter tools, write/spend-gated tools, owner-sensitive tools, and review checklist. Use that packet as the operator contract for ChatGPT setup; do not mark the connector customer-ready until OAuth, packaging, approval, and desktop/mobile smoke tests pass.

The launch packet separates deterministic no-spend read tools from ask_report, which can answer report questions under reports:read. Use get_report_brief first when a stored summary is enough, then use ask_report when the investor explicitly wants a written synthesis.

The connection kit also includes investor workflow packs so the assistant starts with a job, not a blank chat. Current packs cover deal analysis, market watching, campaign operation, and buyer-packet building, each with recommended scopes, starter tools, copyable prompts, and guardrails around owner fields, provider-bound data, and report-credit spend.

Public assistant tools can run redacted parcel previews and listing-aware teaser analysis without paid provider calls. User-scoped tools can list recent reports, pull a source-aware report brief, list the personalized opportunity queue, find thesis-similar opportunities from stored intelligence, create and list area watches, summarize watched-parcel signals, and queue full report jobs from a parcel or attached listing with an idempotency key.

The list_data_coverage tool gives assistant clients a sanitized source map, so AI answers can explain which product lanes are implemented, mixed, provider-required, or planned without exposing internal provider-contract notes.

Assistant clients can also call get_sidecar_package to discover the browser-sidecar manifest, supported marketplaces, data policy, static package files, and listing-handoff endpoints.

Report tools can answer scoped questions against the stored report brief, fetch the personalized synthesis, buyer-confidence packet, visual asset plan, and approved visual analysis directly, and generate or retrieve PDF, DOCX, and Markdown artifacts. Opportunity responses include the investor-fit score, underwriting summary, sanitized listing context, unpriced edge, risks, and next best action. Listing URLs are not returned through MCP opportunity payloads; assistant clients receive marketplace, host, asking price, and price-per-acre context instead.

Campaign tools let assistant clients build a stored-opportunity export plan with ready, review, and suppress queues, estimated owner-record and skip-trace spend, DNC estimate, scrub reasons, CSV/Markdown artifacts, and export columns without buying new provider data.

Comparison tools let assistant clients rank multiple reports using stored investment scores, strategy fit, risk, underwriting, listing context, and source coverage without rerunning paid providers.

Similar-opportunity search lets an assistant ask for deals matching a natural-language buy box, strategy, or cash-flow thesis. It ranks stored opportunities only, reports the matched terms and underwriting signals, and adds no new data charges.

Watch tools let assistant clients list monitored parcels, add an owned or unlocked report to the watched-parcel ledger, and pause, resume, archive, or retune watch criteria. Watch-signal summaries rank stored update events, due reviews, underwriting fit, saved offer criteria, current investor fit, and opportunity scores into proactive why-now cards without new provider or AI calls.

The protected proactive digest route at /api/openland/proactive-digests can email those why-now cards to investors from stored watch intelligence only. Vercel schedules it after market and area-watch refreshes, and the route requires CRON_SECRET plus the normal Resend sender configuration. Targeted dry runs are supported before sending.

Area-watch tools let assistant clients save a county, state, ZIP, or custom thesis, then rank stored opportunities against that market monitor. This is the assistant-native version of a Chrome extension watchlist: the investor can ask for new matches, thesis fit, underwriting fit, and hidden edge from inside their AI workspace.

Area-watch briefs also include a stored market score, demand density, strategy-fit score, recurring-revenue density, and a modeled exit price-per-acre dispersion screen across displayed stored matches. These market analytics use the AcreSignal ledger only and do not trigger new provider or AI calls.

Parcel and owner results are probabilistic. A high-confidence marker means the platform found enough evidence to prioritize review; it does not guarantee buildability, access, owner intent, title condition, or investment outcome.

Use owner enrichment, recent sale status, property type, comps, access, flood, slope, and scrub signals as prioritization inputs, not as legal, financial, or guaranteed business advice.

AcreSignal can flag market-timing priority parcels. This is an inferred outreach signal based on market activity, owner context, parcel profile, recent sale activity, and comp depth.

It does not mean AcreSignal has confirmed seller motivation or investment suitability. Use it as a ranking signal for underwriting and outreach.

New users should start with the guided tutorial. Legal and data-use details are available in the Privacy Policy and Terms of Service.