> ## Documentation Index
> Fetch the complete documentation index at: https://docs.xapichina.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Providers and report coverage

> Interpret CARtech, Chaboshi, source links, category counts, and incomplete extractions.

## Supported report sources

The integration extracts supported CARtech text PDFs and Chaboshi public HTML reports, including supported PDF links resolved through Che168/CXC source links. Provider access and layout determine which findings are available.

Scanned documents without supported text, inaccessible links, expired reports, and unsupported layouts can leave a report advertised without parsed findings. Report URLs can require provider access; the API does not guarantee that every report can be opened publicly.

## Read coverage first

| `parse_status` | Interpretation |
| - | - |
| `not_available` | No supported report is available from the listing evidence |
| `parsed` | Supported report data was extracted |
| `partial` | Some data was extracted, but coverage is incomplete |
| `unsupported` | The report format/provider is not supported for extraction |
| `unavailable` | The advertised report could not be retrieved or made available |

Read this alongside `available` and `data_available`. Do not derive a clean-vehicle label from a missing or inaccessible report.

## Categories and extraction counts

Providers can classify checks under their own exterior, chassis, electrical, powertrain, safety, accident, flood, fire, mileage, and other categories. The API keeps those categories instead of forcing every provider into one inspection checklist.

* `reported_counts`: totals declared by the source report.
* `parsed_item_count`: items actually extracted.
* `count_matches`: whether the parsed count agrees with the provider's declared total, when it can be compared.

`count_matches: true` is a coverage comparison. It does not mean every item was physically inspectable or that the car has no defects. Use `view=full` to inspect statuses and remarks for the individual items.

## Grades and conclusions

Provider grading systems differ. Retain `grade.code`, `grade.source_label`, and `grade.source_meaning`; do not equate one provider's letter grade with another provider's category or with a Korean inspection grade.

`conclusions` retains topic-specific labels and original source fields. An abnormal component, an exterior repair observation, and a provider accident classification are different findings.

## Lookup evidence

`history_queries` can describe maintenance and insurance-event queries, including dates, receipts, and source notices. This is lookup metadata rather than a full service or claim-history dataset. `no_records_returned` is not proof that no prior event occurred.

Use [inspection and insurance evidence](/inspection-and-insurance) for endpoint examples and the distinction between source findings, API-derived summaries, and deprecated compatibility fields.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.