Assembly-accelerated DuckDB MCP server for code search, indexing, and symbol lookup.
Plojure is a playground for Lisp, Python, and x86_64 assembly experiments. The
current hybrid branch also contains a macOS porting path for the FASM
examples.
The target workflow on macOS is intentionally boring:
fasm fasm/basic/fib.asm
./fasm/basic/fib
No Docker command in the hot path, no exec format error, no Linux ELF binary
pretending to be a macOS program.
Install the companion fasm shim from kroq86/fasm-mac:
git clone https://github.com/kroq86/fasm-mac
cd fasm-mac
./install.sh
export PATH="$HOME/.local/bin:$PATH"
The installed command is named fasm.
On Apple Silicon it produces x86_64 Mach-O binaries, so Rosetta is required:
arch -x86_64 /usr/bin/true
If that fails:
softwareupdate --install-rosetta
These examples now use fasm/core/platform.inc instead of hardcoded Linux
syscall numbers:
fasm fasm/basic/fib.asm
./fasm/basic/fib
fasm fasm/basic/arg.asm
./fasm/basic/arg fasm/basic/lol.txt
cd fasm/basic
fasm mycat.asm
./mycat
cd ../advanced/binary_search
fasm bin_s.asm
./bin_s
Expected behavior:
fib.asm prints Fibonacci values.arg.asm prints the file passed as argv[1].mycat.asm prints fasm/basic/lol.txt.bin_s.asm prints the found index.Simple C/Python native-library examples can now build .dylib files on macOS:
cd fasm/advanced/add
./run.sh
cd ../binary_search/wrapper
./run.sh
cd ../../oop_game
./build.sh
The build path is:
fasm --emit=macho-obj add.asm add.o
clang -arch x86_64 -dynamiclib wrapper.c add.o -o mylib.dylib
On Apple Silicon, Python ctypes needs an x86_64/Rosetta Python to load these
libraries. The default Homebrew Python is usually arm64 and cannot load an
x86_64 .dylib.
projects/vector-db now looks for .dylib on Darwin and .so on Linux. The
macOS build path is:
cd projects/vector-db
fasm --emit=macho-obj assembly/dot_product.asm dot_product.o
clang -arch x86_64 -dynamiclib dot_product.o assembly/wrapper.c -o dot_product.dylib
If the native library cannot be loaded, the Python code keeps the existing fallback behavior where available.
format Mach-O formatter..dylib files.The repository also contains a small Lisp interpreter:
python projects/lisp-interpreter/main.py projects/lisp-interpreter/program.lisp
It supports recursive functions, lambda expressions, arithmetic, conditionals, and function calls.
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
docker run -i --rm ghcr.io/kroq86/vector-db-mcp:0.1.1Merge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-kroq86-vector-db-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"ghcr.io/kroq86/vector-db-mcp:0.1.1"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referenceghcr.io/kroq86/vector-db-mcp:0.1.1dockerVector DB MCP works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.