7 releases (4 stable)

1.2.0 Jan 15, 2023
1.1.0 Oct 15, 2022
0.3.0 Aug 28, 2022
0.2.0 Aug 15, 2022
0.1.0 Jul 30, 2022

#300 in Build Utils




build codecov License: GPL v3

Task runner for teams and individuals. Written in Rust.



This project started out of the frustration with existing tools and as a desire of learning Rust. It aims to be easy to use, and for both individuals and teams, specially those working on different platforms. To allow for future improvements in the syntax without breaking the workflow of teams (at least not much, lets be real), it aims to be backward compatible, both between the same mayor release, and with previous mayor releases.

Inspired on different tools like cargo-make, doskey, bash and docker-compose.

Backward compatibility

Starting from version 1.0.0, the goal is to be backward compatible with mayor versions and follow Semantic Versioning. When a mayor version is released, the plan is to support config files from the latest versions by setting the mayor version in the TOML or YAML file with the version key, i.e. version: "1". If the version is not set, the least mayor version supported will be used.


If you have Rust and Cargo installed (rust installation instructions). Then run:

cargo install --force yamis

Pro-tip: make sure ~/.cargo/bin directory is in your PATH environment variable.

Binary releases:

Binaries are also available for Windows, Linux and macOS under releases. To install, download the zip for your system, extract, and copy the binary to the desired location. You will need to ensure the folder that contains the binary is available in the PATH.


When running, if a new version is available, a message will be displayed with the command. The update can be performed by running yamis --update, which will download and replace the binary. Alternatively it can be updated by following the installation instructions again at Installation or Binary releases.

Note that the program will cache the update information for 24 hours, so no need to panic about it performing a request every time you run it.

Quick start

The first step is to add a YAML or TOML file in the project root, i.e. project.yamis.yaml.

Here is a sample YAML file to demonstrate some features:

# project.yamis.yaml
env:  # global env variables
  DOCKER_CONTAINER: sample_docker_container

    private: true   # cannot be invoked directly
      DEBUG: "TRUE"  # Add env variables per task

    help: "Just say hi"  # help message, printend when running `yamis -i say_hi`
    script: "echo Hello {name}"  # takes a name argument, i.e. `--name John`
    script: "echo Hello {name} from Windows"  # Task version for windows systems
  folder_content:  # Default for linux and macOS, can be individually specified like for windows.
    script: "ls {$1?}"  # Takes a single optional argument
    windows:  # Another way of specifying OS specific tasks
      script: "dir {$1?}"

    wd: ""  # Working dir is the dir containing the config file
    program: "docker-compose"
    args: [
      "{$DOCKER_CONTAINER}",  # passes an environment variable into the program arguments
      "{ $@ }"  # passes all extra given arguments 

    bases: ["compose-run", "_debuggable_task"]  # Inherit from other tasks
    args+: ["{$DEBUG?}"]  # Extends args from base task. Here DEBUG is an optional environment variable

After having a config file, you can run a task by calling yamis, the name of the task, and any arguments, i.e. yamis say_hi --name "world". Passing the same argument multiple times will also add it multiple times, i.e. yamis say_hi --name "person 1" --name="person 2" is equivalent to echo Hello person 1 person 2

Note about YAML and TOML files:

The parser used to process YAML files treats values as strings if a string is expected.

For example, the following examples are equivalent

  DEBUG: Yes  # Normally becomes true
  AGENT: 007  # Normally becomes 7
  DEBUG: "Yes"
  AGENT: "007"

However, in the case of TOML files, the parser returns the appropriate type, and therefore it will result in errors.

I.e. the following is not valid

    AGENT = 007

We do not implicitly perform this conversion because we would need to modify the TOML parser. If we performed the conversion after parsing the file we would get AGENT=7 which might be undesired.


Command line options

You can see some help about the command line options by running yamis -h or yamis --help. Essentially, the usage would be like this:

Usage: yamis [OPTIONS] [COMMAND]

  -l, --list              Lists configuration files that can be reached from the current directory
  -t, --list-tasks        Lists tasks
  -i, --task-info <TASK>  Displays information about the given task
  -f, --file <FILE>       Search for tasks in the given file
      --update            Checks for updates and updates the binary if necessary
  -h, --help              Print help information
  -V, --version           Print version information

You can either call a task directly by passing the name of the task and its arguments, i.e. yamis say_hi --name John, or you can specify the configuration file to use with the -f option, i.e. yamis -f project.yamis.yaml say_hi --name John. Note that the -f option is set before the task name, otherwise it would be interpreted as an argument for the task.

The next sections talks about how task files are auto-discovered.

Task files

The task files must be either a TOML or YAML file with the appropriate extension, i.e. project.yamis.toml, or project.yamis.yml. Note that across this document examples are given in either version, but the conversion between them is straightforward.

When invoking a task, starting in the working directory and continuing to the root directory, the program will look configuration files in a certain order until either a task is found, a project.yamis (either TOML or YAML) task file is found, or there are no more parent folders (reached root directory). The name of these files is case-sensitive in case-sensitive systems, i.e. PROJECT.yamis.toml will not work in linux.

The configuration files (in order of precedence, with extension omitted) are named as following:

  • local.yamis: Should hold private tasks and should not be committed to the repository.
  • yamis: Should be used in sub-folders of a project for tasks specific to that folder and sub-folders.
  • project.yamis: Should hold tasks for the entire project.

If the task is still not found, it will look at ~/.yamis/user.yamis.toml or ~/.yamis/user.yamis.yaml or ~/.yamis/user.yamis.yml for user-wide tasks. This is useful for everyday tasks not related to a specific project.


⚠️Warning: DO NOT PASS SENSITIVE INFORMATION AS PARAMETERS IN SCRIPTS. Scripts are stored in a file in the temporal directory of the system and is the job of the OS to delete it, however it is not guaranteed that that will be the case. So any argument passed will be persisted indefinitely.

The script value inside a task will be executed in the command line (defaults to cmd in Windows and bash in Unix). Scripts can spawn multiple lines, and contain shell built-ins and programs. When passing multiple arguments, they will be expanded by default, the common example would be the "{ $@ }" tag which expands to all the passed arguments.

The generated scripts are stored in the temporal directory, and the filename will be a hash so that if the script was previously called with the same parameters, we can reuse the previous file, essentially working as a cache.

Auto quoting

By default, all passed arguments are quoted (with double quotes). This can be changed at the task or file level by specifying the quote param, which can be either:

  • always: Always quote arguments (default)
  • spaces: Quote arguments if they contain spaces
  • never: Never quote arguments

Although quoting prevents common errors like things breaking because an argument with a space was passed, it might fail in certain edge cases.

Replacing the script runner

By default, the script runner in windows is CMD, and bash in unix systems. To use another program you can set the script_runner option in a task. Additionally, you can set script_runner_args which should be a list of extra arguments to pass to the runner before the generated script, i.e. ["-x"] to run the script in bash in debug mode.

You might also want to override the script_ext (or script_extension) option, which is a string containing the extension for the script file, and can be prepended with a dot or not. For some interpreter the extension does not matter, but for others it does. In windows the extension defaults to cmd, and sh in unix.


# Python script that prints the date and time
script_runner = "python"
script_ext = "py"  # or .py
script = """
from datetime import datetime


If using this feature frequently it would be useful to use inheritance to shorten the task. The above can become:

script_runner = "python"
script_ext = "py"  # or .py
private = true

bases = ["_py_script"]
script = """
from datetime import datetime



The program value inside a task will be executed as a separate process, with the arguments passed on args. Note that each argument can contain at most one tag, that is, {$1}{$2} is not valid. When passing multiple values, they are unpacked into the program arguments, i.e. "{$@}" will result in all arguments passed down to the program.

When using inheritance, the arguments for the base can be extended by using args_extend instead of args. This is useful for adding extra parameters without rewriting them.

Running tasks serially

One obvious option to run tasks one after the other is to create a script, i.e. with the following:

yamis say_hi
yamis say_bye

The other option is to use serial, which should take a list of tasks to run in order, i.e.:

serial = ["say_hi", "say_bye"]

Note that any argument passed will be passed to both tasks equally.

It is possible to execute the same task or end with infinite loops. This is not prevented since it can be bypassed by using a script.

Script vs Program:

Because escaping arguments properly can get really complex quickly, scripts are prone to fail if certain arguments are passed. To prevent classic errors, arguments are quoted by default (see Auto quoting), but this is not completely safe. Also, scripts are saved in the temporal directory, and might be persisted indefinitely.

On the other hand, programs run in their own process with arguments passed directly to it, so there is no need to escape them. These can also be extended more easily, like by extending the arguments. The downside however, is that we cannot execute builtin shell commands such as echo, and we need to define the arguments as a list.

Task arguments in the command line

Arguments for tasks can be either passed as a key-value pair, i.e. --name "John Doe", or as a positional argument, i.e. "John Doe".

Named arguments must start with one or two dashes, followed by an ascii alpha character or underscore, followed by any number of letters, digits, - or _. The value will be either the next argument or the value after the equals sign, i.e. --name "John Doe", --name-person1="John Doe", -name_person1 John are all valid. Note that "--name John" is not a named argument because it is surrounded by quotes and contains a space, however "--name=John" is valid named argument.

Named arguments are also treated as positional arguments, i.e. if --name John --surname=Doe is passed, $1 will be --name, $2 will be John, and $3 will be --surname="Doe". Thus, it is recommended to pass positional arguments first.

In you want to pass the arguments as they are to a program, it doesn't matter how they are formatted, you can use the {$@} tag, which will expand to all the arguments.

You can read more about the usage of arguments in tasks in the Tags and Expressions sections.


Tags are used to insert dynamic values into the scripts and arguments of program we want to call. Tags can be used to insert positional and named arguments, environment variables (with a cross-platform syntax) and invoke functions.

The expressions inside tags (including functions) can return either a string, or a list of strings. These are in fact the only two data types that can be used directly in tags. Note that empty lists and lists with a single string will not be coerced into a string to avoid ambiguity, check the Expressions section for more info.

The integer is only allowed when slicing, i.e. {values[0]} is valid, but {1} is not.

In the case of optional expressions, there is no null value, they will simply return an empty string/list. For example {$1?} will return an empty string if $1 is not passed.

Why aren't more data types supported, like integers? Because parsing makes sense only for returning the body of a script or the arguments for a program, and both are always strings or list of strings. Furthermore, it would make more sense to call an external script for more complex operations.


Positional parameters

1-indexed, start with $ and followed by a number, i.e. {$1}, {$2}. These return a single string, so slices of them will return a substring.

Named parameters

Case-sensitive and passed by name, i.e. {out}, {file}, etc. Note that any dash before the argument is removed, i.e. if --file out.txt is passed, {file} will accept it. You can see the Task arguments in the command line section for more info.

These will always return a list of strings, so an index slice will return a string, while a range slice will return a subarray. I.e. { file[0][0] } returns the first character of the first passed file argument, while file[0] will return the first file argument.

All parameters

With { $@ } a list of all arguments will be passed as they are. I.e. if calling a tasks with arguments hello -o file.txt -o=file2.txt, it will return a list with ["hello", "-o", "file.txt", "-o=file2.txt"]. They can be accessed by index and sliced, i.e. { $@[0] } and { $@[0..2] } are valid. Can also be optional, i.e. { $@? }.

Environment variables

Prefixed with $, i.e. { $HOME }, { $PATH }, etc. These are represented as strings, so slices will return a substring. Do not confuse with positional arguments, which are numeric, i.e. $1, or with the all parameters syntax $@.

Note that with this syntax environment variables are loaded when the script or program arguments are parsed, unlike the native syntax that will not work in arguments of programs, and in the case of scripts, will be loaded by the shell. This is intentional to avoid ambiguities and can keep them separate, or use env variables with a different interpreter like python.

String parameters

Strings are another type of valid expressions, but they are more relevant in the function's context. Strings are defined by single or double quotes, cannot contain unescaped new lines. I.e. { "\"hello\" \n 'world'" } is a valid string. Strings can also be sliced, but this is side effect of trying to keep the parser simple rather than a useful feature.

Format strings

These are just regular strings that are treated specially in some functions. I.e. fmt takes a format string and multiple arguments. Each %s occurrence in the string will be replaced with an argument of the same index. Note that in format strings % needs to be escaped with another %, i.e. %%s will be replaced with %s.

For example { fmt("hello %s", $1) } will return hello <first argument>.


For a list of available functions, check the Functions section.

Predefined functions can be used to transform arguments in different ways. They can take values and can be nested. I.e. { join(" ", split(",", $1)) } will split the first argument by ",", and join them back with a space.

At the moment it is not possible to define custom functions as this would require either using an external language such as python, an embedded language such as lua, or implementing a new programming language. One of the goals of this program is to have a simple and clear syntax, so adding support for defining functions breaks this. In most cases where complex operations need to be performed, it would be better and cleaner to have a separate script (i.e. bash or python) that performs the desired operation and then call it from a task with the appropriate arguments. Still, new functions might be added in the future to support flexible argument parsing operations. Feel free to request a new function by submitting a new issue in the repo.

Optional expressions

By default, expressions must return a non-empty string or non-empty array of strings, otherwise an error will be raised. Expressions can be made optional by adding ?, i.e. { $1? }, { map("hello %s", person?)? }, { $@? }, { output? }.

Index and slice

Expressions, including the output of functions can be sliced for more flexibility. The slices are 0 indexed, and accept positive and negative indexes. The whole expression can be either mandatory or optional, i.e. exp[1][0]? does not fail and returns nothing if exp is not set or exp[1] is out of bounds, note that something like exp?[1]?[0]? is invalid.

Here are some examples for parameters hello world -p=1 -p=2 -p=3:

Expression Result
echo { $@[0] } echo hello
echo { $@[0][0] } echo h
echo { p[0] } echo 1
echo { p[:2] } echo 1 2
echo { p[1:] } echo 2 3
echo { $@[0:999] } echo hello world --p=1 --p=2 --p=3
echo { $@[:-1] } echo --p=3
echo { $@[-3:-1] } echo --p=1 --p=2
echo { $@[990:999][0]? } echo
echo { $@[-999]? } echo


Expressions that return an array will be unpacked. For example, given the following tasks:

script = "echo hello {person}"

program = "imaginary-program"
args = ["{ map('-o %s', f) }"]  # map returns an array of strings

If we call yamis hello --person John1 --person John2, it will run echo hello John1 John2. Similarly, yamis something -f out1.txt -f out2.txt will call imaginary-program with ["-o", "out1.txt", "-o", "out2.txt""] parameters. Note that in the last case we call a function called map.

Setting environment variables

Environment variables can be defined at the task level. These two forms are equivalent:

env = {"DEBUG" = "TRUE"}


They can also be passed globally


Also, an env file can be specified at the task or global level. The path will be relative to the config file unless it is an absolute path.

env_file = ".env"

env_file = ".env_2"

If both env_file and env options are set at the same level, both will be loaded, if there are duplicate keys, env will take precedence. Similarly, the global env variables and env file will be loaded at the task level even if these options are also set there, with the env variables defined on the task taking precedence over the global ones.

OS specific tasks

You can have a different OS version for each task. If a task for the current OS is not found, it will fall back to the non os-specific task if it exists. I.e.

  ls: # Runs if not in windows 
    script: "ls {$@?}"

  windows:  # Other options are linux and macOS
    script: "dir {$@?}"

Os tasks can also be specified in a single key, i.e. the following is equivalent to the example above.

    script: "ls {$@?}"

    script: "dir {$@?}"

Note that os-specific tasks do not inherit from the non-os specific task implicitly, if you want to do so, you will have to define bases explicitly, i.e.

      DIR: "."
    script: "ls {$DIR}"

    bases: [ls]
    script: "dir {$DIR}"

Working directory

By default, the working directory of the task is one where it was executed. This can be changed at the task level or root level, with wd. The path can be relative or absolute, with relative paths being resolved against the configuration file and not the directory where the task was executed, this means "" can be used to make the working directory the same one as the directory for the configuration file.

Documenting tasks

Tasks can be documented using the help key. Unlike comments, help will be printed when running yamis -i <TASK>. Note that help is inherited. If you wish to remove it, you can set it to "".

Task inheritance

A task can inherit from multiple tasks by adding a bases property, which should be a list names of tasks in the same file. This works like class inheritance in common languages like Python, but not all values are inherited.

The inherited values are:

  • wd
  • help
  • quote
  • script
  • script_runner
  • script_runner_args
  • script_ext
  • script_extension (alias for script_ext)
  • program
  • args
  • serial
  • env (the values are merged instead of overwriting)
  • env_file (the values are merged instead of overwriting)

Values not inherited are:

  • args_extend (added to the inherited args and destroyed afterwards)
  • args+ (alias for args_extend)
  • private

The inheritance works from bottom to top, with childs being processed before the parents. Circular dependencies are not allowed and will result in an error.

It will attempt to find and the os-specific task first and inherit from it, if not found, it will use the regular task. For example:

    script: "echo hello"
    script: "echo hi"
    bases: [sample]

Is equivalent to:

    script: "echo hello windows"
    script: "echo unix"
   script: "echo hello windows"

This way base tasks can be defined for each OS, and have only one version for its children. However, note that os-specific tasks do not inherit implicitly from the non os-specif task. As in the above example, sample.windows will not implicitly inherit from sample.

Extending program arguments

Args can be extended with args_extend or it's alias args+. These will append the given list to the args inherited from the bases.


    program: "program"
    args: ["{name}"]

    bases: ["program"]
    args_extend: ["{phone}"]

    env: {"KEY": "VAL"}
    args: ["{other_param}"]
    private: true  # cannot be called directly, field not inherited

    bases: ["program_extend", "other"]
    args+: ["{address}"]  # args+ is an alias for args_extend

In the example above, program_extend_again will be equivalent to

    program: "program"
    env: {"KEY": "VAL"}
    args: ["{name}", "{phone}", "{address}"]

Private tasks

Tasks can be marked as private by setting private = true. Private tasks cannot be called by the user, but are useful for inheritance.

Debug Options

Some debug options can be added at the task or file level under debug_config

  • print_file_path: Boolean, defined at the file level and false by default. If true, the absolute config file path will be displayed when running a task
  • print_task_name: Boolean, defined at the task or file level, true by default. If true, the name of the task will be displayed when tunning a task

List of functions

List of predefined functions.

map function

Signature: map<S: str | str[]>(fmt_string: str, values: S) -> S

Maps each value to fmt(fmt_string, val).



  quote: never
  script: |
    echo {map("'%s'", $@)}

  program: merge_txt_files
  args: ["{map('%s.txt', $@)}"]

yamis sample person1 person2 will result in echo hi 'person1' 'person2'

yamis sample2 file1 file2 will result in calling merge_txt_files with arguments ["file1.txt", "file2.txt"]

join function

Signature: join<S: str | str[]>(join_str: str, values: S) -> str

The first parameter of join is a string that will be inserted between all values given in the second parameter returning a single string. If the second parameter is a single string, it will be returned as is.


  • join_str: String to insert between the values
  • values: Value or values to join


  quote: never
  script: |
    echo hello {join(" and ", $@)}

yamis sample person1 person2 will result in echo hi person1 and person2'

jmap function

Signature: jmap<S: str | str[]>(fmt_string: str, values: S) -> S

Shortcut for join("", map(fmt_string, values))



  quote: never
  script: |
    echo hi{jmap(" '%s'", $@)}

  program: some_program
  args: ["{jmap('%s,', $@)}"]

yamis sample person1 person2 will result in echo hi 'person1' 'person2'

yamis sample2 arg1 arg2 will result in calling some_program with arguments ["arg1,arg2,"]

fmt function

Signature: fmt(fmt_string: str, *args: str) -> str

The first parameter of fmt is a format string, and the rest of the values are parameters to format the string with. Note that those extra parameters must be string values, not list of strings, i.e. cannot pass directly $@.


  • fmt_string: format string
  • args: Arguments that will replace the %s occurrence of the same index


  quote: never
  script: |
    echo {fmt("Hi %s and %s", $1, $2)}

yamis sample person1 person2 will result in echo Hi person1 and person2

trim function

Signature: trim<S: str | str[]>(value: S) -> S

Removes leading and trailing whitespaces (including newlines) from the string or each string in list of strings.


  • value: String or list of strings to trim


  quote: never
  script: |
    echo {trim("  \n  hello world  \n")}

yamis sample will result in echo hello world

split function

Signature: split(split_val: str, split_string: str) -> str

Splits the string with the given value


  • split_val: Value to split by
  • split_string: String to split


  quote: never
  script: |
    echo {split(",", "a,b,c")}

yamis sample will result in echo a b c


Why not use make, cargo-make or cmake and similar tools?

They are great tool, but sometimes simpler is better. I created this tool because I didn't feel comfortable existing solutions, specially on the arguments parsing side, but you can use whatever fits your needs, no hard feelings. Also, there should be no any issue with using this together with other tools.

Can I define my own functions?

Not possible at the moment, unless you fork the repository and add your own (which is easy to do if you know rust). You can always contribute so everyone benefits from it. While I have plans to add more functions in the future, allowing custom functions is not that straightforward and probably not worth the effort, perhaps it is better to use a separate script, i.e. python. I am open to suggestions.


Feel free to create issues to report bugs, ask questions or request changes.

You can also fork the repository to make pull requests, just make sure the code is well tested. Signed commits are preferred.


~640K SLoC