fp-conv MCP
An MCP server providing AI agents with local tools to encode, decode, and convert numbers across various floating-point formats including FP32, FP16, BF16, TF32, FP8, FP6, FP4, and integer types. It supports rounding modes, overflow saturation behavior, and precision loss calculation.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Encode decimal values into specific binary formats like FP16 or FP8
- Decode binary or hex bit-patterns back to decimal values
- Convert values between different numeric formats with specified rounding
- List available format presets and get detailed format information
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
fp-conv-mcp 1.5.1
{
"version": "1.5.1",
"transport": "stdio",
"identifier": "fp-conv-mcp",
"runtimeHint": "npx",
"registryType": "npm"
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"fp-conv-mcp": {
"command": "npx",
"args": [
"-y",
"fp-conv-mcp@1.5.1"
]
}
}
}Cline
{
"mcpServers": {
"fp-conv-mcp": {
"command": "npx",
"args": [
"-y",
"fp-conv-mcp@1.5.1"
]
}
}
}VS Code
{
"servers": {
"fp-conv-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"fp-conv-mcp@1.5.1"
]
}
}
}Claude Desktop
{
"mcpServers": {
"fp-conv-mcp": {
"command": "npx",
"args": [
"-y",
"fp-conv-mcp@1.5.1"
]
}
}
}Let your agent find its tools
Find servers like fp-conv MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.