# GroundTruth — subsurface scan QA & trade estimating [Health: Active]

**Category:** 🏭 Industrial & IoT  
**Repository:** https://gtfi.ai/for-agents  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/groundtruth-subsurface-scan-qa-trade-estimating

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

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What GroundTruth — subsurface scan QA & trade estimating does

GroundTruth — subsurface scan QA & trade estimating is described as an MCP server for reviewing ground-penetrating radar (GPR) scans. Its stated workflow uses two independent AI models to read the scan, identify locate or pipe scan findings, and flag a do-not-core call.

The main value for an agent workflow is focused scan interpretation rather than general-purpose construction management. A calling agent can use the server when GPR information needs to be reviewed before a coring decision, or when scan findings need to inform subsurface trade estimating.

## Tools and capabilities

- Reading GPR scans with two independent AI models.
- Identifying locate or pipe scan findings.
- Flagging a do-not-core call.
- Supporting subsurface scan QA and trade-estimating workflows, as indicated by the listing name.

No individual MCP tool names are provided. The source also does not document operations for uploading scans, retrieving prior results, exporting reports, or editing findings. Those details should be confirmed before implementation.

## How it works

The described process is model-based and uses two separate AI model passes over GPR data. Their stated purpose is to read the scan, locate or identify pipe-related scan information, and mark a do-not-core decision when applicable. GroundTruth — subsurface scan QA & trade estimating therefore fits a workflow where an agent supplies or makes available GPR scan information and uses the resulting interpretation as part of a review.

The material does not state whether the models run remotely, whether they compare their answers, or whether a human must approve the result. It also does not define the supported scan file formats, image or data requirements, response fields, or error behavior.

## Limitations and notes

There is no README, and the source description does not identify an installation command, package name, endpoint, authentication method, environment variables, MCP transport, or supported client applications.

The listing establishes scan reading and do-not-core flagging, but it does not substantiate broader estimating features such as quantity takeoffs, pricing, bid generation, project-system integrations, or cost databases. It should be evaluated specifically for the documented GPR review use case until those capabilities are confirmed. Compatibility with a particular MCP client cannot be inferred from the available information.

_Full upstream README: https://allmcps.com/mcp/groundtruth-subsurface-scan-qa-trade-estimating/readme_