Hachitool

Hachitool is a set of utilities that make it easier to work with Python scripts in GitHub Actions.

Installation

Hachitool can be installed like any other Python package:

- run: pip install hachitool

Inline scripts

Hachitool can be ephemerally installed for inline Python scripts via uv:

- uses: astral-sh/setup-uv@v3

- shell: uv run --with hachitool python {0}
  run: |
    import hachitool

    # Do stuff here

External scripts

Hachitool can be emphemerally installed for external scripts via uv and inline script metadata:

# script.py

# /// script
# dependencies = [
#     "hachitool",
# ]
# ///

import hachitool

# Do stuff here

# workflow.yml

- uses: astral-sh/setup-uv@v3

- run: uv run script.py

Usage

hachitool.set_output

Set output parameters for a step. Takes either:

• a key as its first argument and a value as its second
• a set of key-value pairs as either a dictionary or keyword arguments

import hachitool

# All of these are equivalent
hachitool.set_output("key", "value")
hachitool.set_output({"key": "value"})
hachitool.set_output(key="value")

hachitool.set_env

Set environment variables. Takes either:

• a key as its first argument and a value as its second
• a set of key-value pairs as either a dictionary or keyword arguments

import hachitool

# All of these are equivalent
hachitool.set_env("key", "value")
hachitool.set_env({"key": "value"})
hachitool.set_env(key="value")

hachitool.add_path

Append something to the system path.

import hachitool

hachitool.add_path("/absolute/or/relative/path")

hachitool.summary.add

Add content to the step summary. Supports GitHub-flavored Markdown.

import hachitool

hachitool.summary.add("this is a summary")

Tip

Calling hachitool.summary directly does the same thing:

import hachitool
   
hachitool.summary("this is a summary")

If the keyword-only overwrite argument is True, existing summary content will be erased:

import hachitool

hachitool.summary.add("this is a summary", overwrite=True)

hachitool.summary.clear

Clear the step summary.

import hachitool

hachitool.summary.clear()

hachitool.mask

Mask a value.

import hachitool

hachitool.mask("super secret value")

hachitool.log

Print a message to the log. Takes the following arguments:

Argument	Type	Description	Required?
level	"debug" | "notice" | "warning" | "error"	The log level of the message.	Yes
message	str	The message to print.	Yes
file	str	The path to a file to annotate with the message.	No
line	int | tuple[int, int]	The line(s) of file to annotate with the message. A tuple will be interpreted as a pair of starting and ending lines.	No
column	int | tuple[int, int]	The column(s) of file to annotate with the message. A tuple will be interpreted as a pair of starting and ending columns.	No

level and message are the first and second positional arguments, respectively. file, line, and column are keyword-only.

import hachitool

hachitool.log("notice", "this is a notice message", file="main.py", line=1, column=6)

# Using tuples for `line` and `column`
hachitool.log("notice", "this is a notic message", file="main.py", line=(1, 5), column=(6, 10))

hachitool.debug, hachitool.notice, hachitool.warning, hachitool.error

Print a debug, notice, warning, or error message to the log, respectively. Takes all arguments of hachitool.log except for level.

import hachitool

hachitool.debug("this is a debug message")
hachitool.notice("this is a notice message")
hachitool.warning("this is a warning message")
hachitool.error("this is an error message")

hachitool.fail

Optionally print an error-level message, then fail the workflow. Takes an optional exit_code argument that must be an integer greater than or equal to 1. Additionally takes all arguments of hachitool.error, except message is optional.

import hachitool

hachitool.fail("something went wrong", exit_code=1)

hachitool.log_group

Anything printed to the log inside this context manager will be nested inside an expandable group.

Takes a mandatory title argument.

import hachitool

with hachitool.log_group(title="group title"):
    print("I'm part of a log group!")
    print("me too!")
    print("me three!")

hachitool.literal

Nothing printed to the log inside this context manager will be interpreted as a workflow command.

import hachitool

hachitool.warning("this is a warning message to show that commands are being processed")

with hachitool.literal():
    hachitool.warning("this will not render as a warning because commands are not being processed")

hachitool.warning("this will render as a warning since commands are being processed again")