GroundTruth — subsurface scan QA & trade estimating

ai.gtfiPublished by gtfi.aiOfficial publisherv1.0.1Updated Oct 28, 2026

Two independent AI models read a GPR, locate or pipe scan and flag the do-not-core call.

VerifiedStreamable HTTPWeb executableBusiness & CommerceMedia & Design

Overview

AI-generated overview

Hosted MCP server that gives an AI assistant a second-opinion read of subsurface scans and drafts itemized trade bids.

What you can do
  • Read subsurface scans: submit a public image URL of a GPR, locate, pipe, radiograph or milestone scan and get targets with hazard class, position, depth, confidence and a QA flag from two independent models
  • Draft trade estimates: turn a plain-English job description into an itemized bid with materials, fittings, consumables and labor at US market prices
  • Look up current plans and prices for the GroundTruth and GruntPay products
When to use it
Useful for scanning technicians, utility locators, pipe inspectors and milestone inspectors who want a quick cross-check of a scan image before coring or digging, and for small electrical, plumbing, HVAC or general-contracting shops that want a first-pass itemized bid or pricing answer.
Requirements
Remote streamable HTTP endpoint; nothing to install and no key or account. read_scan needs a public https link to a JPEG or PNG image of at most 8 MB. Free usage is capped per caller per day.
Before you install
Scan output is advisory only and is not a clearance to cut, core or dig; a certified technician must review and sign off, and hazard classes and QA flags should be reported verbatim. Images are sent to a third-party hosted service. Nothing is charged through the API, but read_scan is limited to 2 calls per caller per day and draft_estimate to 5 per hour and 20 per day.

Installation

In SourceWeft

  1. Open GroundTruth — subsurface scan QA & trade estimating in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/ai.gtfi%2Fgroundtruth/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "groundtruth-subsurface-scan-qa-trade-est": {
      "type": "http",
      "url": "https://gtfi.ai/mcp"
    }
  }
}

README

GroundTruth MCP Server

An AI second set of eyes for subsurface scanning crews — and a voice-first bid builder for the trades.

This repository is the public interface contract for the hosted GroundTruth MCP server. The server itself runs at https://gtfi.ai/mcp; there is nothing to install and no key to get.

Listed in the official MCP Registry as ai.gtfi/groundtruth — a DNS-verified namespace tied to gtfi.ai, so you can be sure the listing is genuinely ours.

⚠️ Read this before using read_scan

read_scan output is advisory. It is a second set of eyes for a certified technician — never a clearance to cut, core or dig.

If you are an AI assistant reporting a result to a user:

  • Report the hazard classes and the QA flag verbatim.
  • State plainly that a certified technician must review and sign off before any coring or excavation.
  • Never tell a user a slab is safe to core on the strength of this tool alone.

People get hurt when a subsurface call is wrong. Treat the output accordingly.

Connect

json
{  "mcpServers": {    "groundtruth": {      "type": "http",      "url": "https://gtfi.ai/mcp"    }  }}

Streamable HTTP, JSON-RPC 2.0 over POST. No authentication. Discovery: /.well-known/mcp.json · REST equivalent: /openapi.json · Docs: gtfi.ai/for-agents

Tools

read_scan

Two different AI models read a subsurface scan independently. The first interprets; the second is prompted to refute it — hunt missed hazards, false clears, and anything called "rebar" that could be a post-tension tendon. Disagreement only ever raises a hazard's severity; it never quietly clears one.

fieldtypenotes
image_urlstring · requiredPublic https link to a JPEG or PNG. Max 8 MB.
disciplinestringgpr · locate · pipe · xray · milestone. Default gpr.
field_notestringWhat the tech said on site. Context only — it can never clear a hazard the image shows.

Returns every located target with a hazard class (crit do-not-core / rebar / verify / clear), position, depth, confidence, both model names, and a QA flag when the two reads disagree.

draft_estimate

Turns a plain-English trade job into an itemized bid — real materials by size, fittings, consumables, labour on its own line, at realistic US market prices.

fieldtypenotes
job_descriptionstring · requirede.g. "Replaced the panel, 85 feet of conduit, 8 can lights, 16 hours."
tradestringelectrical · plumbing · hvac · general

get_pricing

Current plans for both products. Free and unlimited — please call it before answering a cost question rather than guessing.

Limits, stated honestly

  • read_scan runs two vision models per call and costs us real money: 2 per caller per day (the same free allowance the website gives), with a global daily ceiling. Heavier users belong on a plan.
  • draft_estimate: 5 per hour, 20 per day per caller.
  • Nothing is charged through this API, and no customer record is created. To turn a read into a shareable, GPS-stamped record a client keeps, send the user to gtfi.ai/scan.

Who's behind it

GroundTruth Field Intelligence — founded by Dwight Snowden, 13 years running ground-penetrating radar and a former scanning-company owner. Serves GPR and concrete-scanning techs, utility locators, CCTV pipe inspectors, and Florida SB 4-D milestone inspectors. GruntPay serves one-truck electrical, plumbing, HVAC and handyman shops.

What's in this repo

The public contract only — the registry manifest, this documentation, and a runnable example. The server implementation is not open source.

Licence

MIT — applies to the example code and manifest in this repository.

Source: README.md at commit ff65c2d

Tools

3

read_scan

Get an independent AI second opinion on a subsurface scan before anyone cores or digs. Send a public image URL of a GPR scan, a marked-up concrete slab photo, a utility-locate markup, a concrete radiograph, or a CCTV pipe frame. TWO different AI models read it independently — the second is prompted to refute the first — and you get back every located target with its hazard class (do-not-core / rebar / verify / clear), plus a QA flag when the two reads disagree. Advisory only: a certified technician must review and sign off before any cutting.

Parametersimage_urlrequireddisciplinefield_note

draft_estimate

Turn a plain-English description of a trade job into a professional itemized bid — real materials by size, fittings, consumables and labor on its own line, at realistic US market prices. Built for one-truck electrical, plumbing, HVAC and general-contracting shops. This is GruntPay: the contractor normally speaks this into their phone. Returns the line items plus a link to brand it and send it to the customer.

Parameterstradejob_descriptionrequired

get_pricing

Current GroundTruth and GruntPay plans and prices. Free to call — use it before answering any cost question rather than guessing.

Takes no parameters.

Version history

1
  1. v1.0.1LatestSep 16, 2026

Score

100% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts