feat: Drop remaining references to Docker and containers

README.md

@@ -7,11 +7,9 @@
 
 Leverage CLI is the tool used to manage and interact with any Leverage project.
 
-It transparently handles the most complex and error prone tasks that arise from working with a state-of-the-art
-infrastructure definition like our Leverage Reference Architecture. Leverage CLI uses a dockerized approach to
-encapsulate the tools needed to perform such tasks and to free the user from having to deal with the configuration and
-management of said tools.
-Provides the means to interact with your Leverage project and allows you to define custom tasks to run.
+It transparently handles the most complex and error-prone tasks that arise from working with a state-of-the-art infrastructure definition like our Leverage Reference Architecture. It understands the structure of the Leverage Reference Architecture and assists the user in their day-to-day operation.
+
+Leverage CLI orchestrates a set of tools used to operate in a Leverage project. Usually this means OpenTofu/Terraform, but also aws-cli, kubectl and more. 
 
 Reviewing and implementing the [Binbash Leverage Landing Zone for AWS](https://leverage.binbash.co/try-leverage/) would
 be a very good place to start!
@@ -23,10 +21,13 @@ to [this page](https://leverage.binbash.co/user-guide/leverage-cli/installation/
 
 ### Note for migration from previous versions
 
-If you come from Leverage CLI version <1.8.0 and want to install Leverage CLI version >= 1.8.0 keep into account the
+If you come from Leverage CLI version <3.0.0 and want to use Leverage CLI version >= 3.0.0 keep into account the
 following.
 
-The `build.env` file format has changed. As an example, this is the old format:
+Leverage CLI no longer runs tools inside a Docker container; it executes the OpenTofu/Terraform binaries installed on
+your host. Because of this, the `TERRAFORM_IMAGE_NAME`/`TF_IMAGE_NAME` and `TERRAFORM_IMAGE_TAG`/`TF_IMAGE_TAG`
+parameters in the `build.env` file no longer exist. In their place, the optional `TF_BINARY` parameter specifies the path
+to the OpenTofu/Terraform binary to use. As an example, this is the old format:
 
 ```
 # Project settings
@@ -36,8 +37,7 @@ PROJECT=bb
 MFA_ENABLED=false
 
 # Terraform
-TERRAFORM_IMAGE_NAME=binbash/terraform-awscli-slim
-TERRAFORM_IMAGE_TAG=1.1.9
+TF_IMAGE_TAG=1.5.0-0.2.0
 ```
 
 New version example:
@@ -49,16 +49,19 @@ PROJECT=bb
 # General
 MFA_ENABLED=false
 
-# Terraform
-TF_IMAGE_TAG=1.5.0-0.2.0
+# OpenTofu/Terraform binary
+TF_BINARY=/usr/local/bin/tofu
 ```
 
-So, if you have created a project with version <1.8.0 and want to use it with version >=1.8.0 you should:
-
-- remove TERRAFORM_IMAGE_NAME line
-- update TF_IMAGE_TAG from this form '9.9.9' to this one '9.9.9-9.9.9'.
+So, if you have created a project with version <3.0.0 and want to use it with version >=3.0.0 you should:
 
-For the second item you can check the version [here](https://hub.docker.com/r/binbash/leverage-toolbox/tags).
+- remove the `TERRAFORM_IMAGE_NAME`/`TF_IMAGE_NAME` and `TERRAFORM_IMAGE_TAG`/`TF_IMAGE_TAG` lines
+- install [OpenTofu](https://opentofu.org/docs/intro/install/) or [Terraform](https://developer.hashicorp.com/terraform/install)
+  on your host
+- optionally, set `TF_BINARY` to the path of the binary. This is useful when you have multiple versions of
+  OpenTofu/Terraform installed, or when the binary lives in an unconventional location. It can be an absolute path, a path starting with `~`, or a path
+  relative to the project root. If not set, the `tofu` or `terraform` binary is looked up in your `PATH`. Keep in mind
+  the binary must match the command you use: `leverage tofu` expects OpenTofu and `leverage terraform` expects Terraform.
 
 ## System requirements
 

leverage/logger.py

@@ -174,7 +174,7 @@ def get_tasks_logger():
 
 def _raw_logger():
     """
-    Provide a raw logger, in case we need to print stuff that already comes formatted (like some container logs).
+    Provide a raw logger, in case we need to print stuff that already comes formatted.
     """
     logger = logging.getLogger("raw")
     logger.setLevel(logging.INFO)

leverage/modules/credentials.py

@@ -253,7 +253,6 @@ def credentials(state):
             raise an exception
 
     If we reached the only common.tfvars scenario, we have no project name nor TF_IMAGE_TAG.
-    So the best chance is to read the common.tfvars directly without a container, e.g. with sed or grep
     """
     project_config = _load_project_yaml()
     build_env = Path(f"{PROJECT_ROOT}/build.env")

leverage/modules/tf.py

@@ -23,8 +23,8 @@
 @pass_state
 def tofu(state):
     """Run OpenTofu commands in the context of the current project.
-    All tofu subcommands that receive extra args will pass the given strings as is to their corresponding OpenTofu
-    counterparts in the container. For example as in `leverage tofu apply -auto-approve` or
+    All tofu subcommands that receive extra args will pass the given strings as is to the OpenTofu binary.
+    For example as in `leverage tofu apply -auto-approve` or
     `leverage tofu init -reconfigure`
     """
     state.runner = TFRunner(binary=state.paths.tf_binary, env_vars=state.environment)
@@ -34,8 +34,8 @@ def tofu(state):
 @pass_state
 def terraform(state):
     """Run Terraform commands in the context of the current project.
-    All terraform subcommands that receive extra args will pass the given strings as is to their corresponding Terraform
-    counterparts in the container. For example as in `leverage terraform apply -auto-approve` or
+    All terraform subcommands that receive extra args will pass the given strings as is to the Terraform binary.
+    For example as in `leverage terraform apply -auto-approve` or
     `leverage terraform init -reconfigure`
     """
     state.runner = TFRunner(binary=state.paths.tf_binary, terraform=True, env_vars=state.environment)

leverage/modules/tfautomv.py

@@ -12,7 +12,7 @@
 @authenticate
 @pass_state
 def tfautomv(state, args):
-    """Run TFAutomv commands in the context of the current project.`"""
+    """Run TFAutomv commands in the context of the current project."""
     tf_default_args_string = " ".join(tf_default_args())
     state.environment["TF_CLI_ARGS_init"] = tf_default_args_string
     state.environment["TF_CLI_ARGS_plan"] = tf_default_args_string