current_timeLABSCO SUMMARY
The small stuff, so you stop writing it each time.
None of these tools is hard to write; having them all already there is the point — hash a file, count words, pull the emails out of a blob, check whether a URL resolves, without a script per task. It reads the local filesystem, environment variables and process list, so run it somewhere you are comfortable with an agent seeing all three.
What it is
A Python server carrying a wide set of small tools, grouped into date and time, local file handling, system information, text processing, and web operations.
What you get
- Date and time: current time, timestamp formatting, age from a birth date, days between two dates, a month calendar
- Files: read a file, detailed file information, list a directory, MD5/SHA256 hashing, pattern search
- System: general system information, CPU, memory, disks, network interfaces, running processes, environment variables
- Text: word and line counts, regex search and replace, email and URL extraction, text analysis, encode/decode, hashing, chunking
- Web: HTTP requests, URL parsing and building, URL encode/decode, format and reachability validation, domain extraction
- Async stdio communication, configurable logging, and a CLI
Requirements
Python 3.10 or later and uv; installed from source. Everything except the web tools acts on the machine it runs on — the environment variables and process list tools read that machine's real environment and processes.
Setup effort
Build from source — git clone https://github.com/SergiFuster/mcp-custom-tools.git
USE CASES
01
Inspecting files without a shell
Read a file, get its details, list the directory around it, and hash it to compare against another copy.
02
Getting structure out of a wall of text
Count words and lines, run a regex replace, and pull out the emails and URLs it contains.
03
Checking what a machine is doing
Read CPU, memory, disk and network information, and the list of running processes.
04
Doing date arithmetic without getting it wrong
Days between two dates, an age from a birth date, a timestamp formatted readably, a month laid out as a calendar.
05
Poking at a URL
Make the HTTP request, parse the URL into parts, check that it validates and resolves, and pull the domain out.
06
Splitting text to fit somewhere else
Chunk a long string, then encode or hash it for whatever consumes it next.
31 tools
Listed in project docs. Your client may show a different list.
format_timestampcalculate_agedays_betweenmonth_calendarread_filefile_infolist_directorycalculate_hashsearch_filessystem_infocpu_infomemory_infodisk_infonetwork_infoprocess_listenvironment_varsword_countsearch_replaceextract_emailsextract_urlstext_analysisencode_decodegenerate_hashsplit_texthttp_requestparse_urlbuild_urlurl_encode_decodevalidate_urlextract_domainBEFORE YOU INSTALL
What to know first
Tools you get
31 tools named in the project's documentation.
Requires
- uv on PATH
- Python 3.10+
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone this one yourself.
Server setup
TERMINAL
git clone https://github.com/SergiFuster/mcp-custom-tools.git
cd mcp-custom-tools
uv sync
uv run mcp-server --log-level INFOMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"mcp-custom-tools": {
"command": "uv",
"args": [
"--directory",
"<YOUR_PATH_TO_MCP_CUSTOM_TOOLS>",
"run",
"mcp-server",
"--log-level",
"INFO"
]
}
}
}
