
Schema Mapper
Maps flat vendor CSV/JSON files to your schema with human review and exact values
Schema Mapper is a freemium MCP server for data quality. It runs locally on your machine and is started by your AI client.
- Pricing
- Freemium
- Platforms
- CLI, API, Linux
- Runs
- Locally
- Version
- 1.0.0
- Listing updated
Similar MCP servers
Install
Run in a terminal
claude mcp add schema-mapper -- npx -y https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpbAdd to claude_desktop_config.json
{
"mcpServers": {
"schema-mapper": {
"command": "npx",
"args": [
"-y",
"https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"schema-mapper": {
"command": "npx",
"args": [
"-y",
"https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"schema-mapper": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"schema-mapper": {
"command": "npx",
"args": [
"-y",
"https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"schema-mapper": {
"command": "npx",
"args": [
"-y",
"https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb"
]
}
}
}Using several servers? Build one config file with the MCP config generator.
About Schema Mapper
Schema Mapper takes a flat vendor export and proposes how its columns line up with a target contract you define. Each proposed match comes with its supporting evidence and any ambiguity, so a named reviewer approves every correspondence and cast before anything is generated.
Once approved, the mapping is validated against all actual records and packaged as a versioned mapping file plus a Python wrapper that imports the free runtime. Identifiers with leading zeros stay as text, and decimal values are kept exact rather than rounded through floats. Structural or format changes in later files are flagged for review instead of passing silently.
The local CLI, Python SDK and stdio MCP server are free under Apache-2.0 and need no hosted key. The hosted option handles preview and validation for free, and charges a one-time $9 for the approved artifact download. It is aimed at data engineers and integration developers who receive recurring vendor files and need a documented, repeatable mapping rather than ad hoc scripts.
What it does
- Evidence-backed correspondence proposals
- Ranks candidate source-to-target field matches and shows the evidence behind each one, with scores treated as rankings rather than probabilities.
- Named review before approval
- Requires a named reviewer to confirm each correspondence and cast before a mapping can be approved.
- Exact identifiers and decimals
- Preserves leading-zero identifiers as text and keeps decimal values exact during transformation.
- Reusable generated transform
- Produces a versioned mapping and a Python wrapper that depends on the matching free runtime.
- Drift detection on reuse
- Stops processing when a later file's structure or format no longer matches the approved mapping.
- Local MCP server
- Exposes inference, review, approval, validation and transformation as nine stdio tools for MCP clients.
Good for
- Map a recurring supplier CSV onto an internal customer import schema
- Keep account codes and product IDs with leading zeros intact during import
- Reuse an approved vendor mapping and get alerted when the export layout changes
- Run the mapping workflow locally from an AI coding agent over stdio MCP
Pricing
Freemium
Local CLI, SDK and MCP server free; hosted preview and validation free; one-time $9 artifact download
Prices change. Check schema.contentsweep.ai before you buy.
Schema Mapper FAQ
What file formats are supported?
Flat CSV, JSON and JSONL files. The hosted service accepts sources up to 1 MiB, while the local runtime allows core files up to 4 MiB.
What does the free version include?
The Apache-2.0 CLI, Python SDK and stdio MCP server are free and need no hosted key. Hosted preview and validation are also free.
What does the $9 purchase include?
A one-time download of the approved artifact: the mapping, the Python transform, a validation receipt and a README. It does not include your source records.
What is required to run the generated transform?
Python 3.12 or later and the matching Schema Mapper runtime. The generated file imports that runtime and is not self-contained.
Does Schema Mapper handle nested data?
No. Nested values, locale or date conversions and arbitrary scripts are listed as unsupported.
How do I add Schema Mapper to Claude Code?
Run this command in your terminal: claude mcp add schema-mapper -- npx -y https://github.com/founder-sweep/contentsweep-schema-mapper/releases/download/mcpb-v1.0.0/contentsweep-schema-mapper-1.0.0.mcpb
What are the best alternatives to Schema Mapper?
The closest alternatives to Schema Mapper are csvconvert.uk and Multipart row preflight.
Where this comes from
- Description: written from the project's README, the project's website and its MCP Registry entry.
- Install steps: from the official MCP Registry entry.