diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
new file mode 100644
index 0000000..61d89f6
--- /dev/null
+++ b/.github/workflows/tests.yml
@@ -0,0 +1,29 @@
+name: tests
+
+on:
+  push:
+    branches: [ main, dev ]
+  pull_request:
+    branches: [ main, dev ]
+
+jobs:
+  tests:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v6
+
+      - name: Parse environment.json and set Docker image
+        id: docker_image
+        run: |
+          IMAGE=$(jq -r '.base_image' .codeocean/environment.json)
+          IMAGE=${IMAGE//codeocean/nciccbr}
+          echo "image=$IMAGE" >> $GITHUB_OUTPUT
+          echo "Using Docker image: $IMAGE"
+
+      - name: Run tests in Docker
+        run: |
+          docker run --rm \
+            -v ${{ github.workspace }}:/workspace \
+            -w /workspace \
+            ${{ steps.docker_image.outputs.image }} \
+            Rscript tests/testthat.R
diff --git a/code/main.R b/code/main.R
index ca06bfa..e9bff11 100644
--- a/code/main.R
+++ b/code/main.R
@@ -1,5 +1,4 @@
 #!/usr/bin/env Rscript
-rlang::global_entrace()
 library(argparse)
 library(glue)
 library(MOSuite)
@@ -13,17 +12,67 @@ setup_capsule_environment()
 # parse CLI arguments
 parser <- ArgumentParser()
 
-parser$add_argument("--count_type", type="character", default="filt")
-parser$add_argument("--sub_count_type", type="character", default=NULL, help="Sub count type if count_type is a list")
-parser$add_argument("--sample_id_colname", type="character", default=NULL, help="Column name for sample IDs")
-parser$add_argument("--feature_id_colname", type="character", default=NULL, help="Column name for feature IDs")
-parser$add_argument("--samples_to_include", type="character", default="", help="Comma-separated list of samples to include")
-parser$add_argument("--covariates_colnames", type="character", default="Group", help="Comma-separated list of covariate column names")
-parser$add_argument("--contrast_colname", type="character", default="Group", help="Column containing group variables for DE")
-parser$add_argument("--contrasts", type="character", default="", help="Contrasts in format group1-group2,group1-group3")
-parser$add_argument("--input_in_log_counts", type="logical", default=FALSE, help="Counts are already log2-transformed")
-parser$add_argument("--return_mean_and_sd", type="logical", default=FALSE, help="Return mean and SD in addition to DE estimates")
-parser$add_argument("--voom_normalization_method", type="character", default="quantile", help="Normalization method for limma::voom")
+parser$add_argument("--count_type", type = "character", default = "filt")
+parser$add_argument(
+  "--sub_count_type",
+  type = "character",
+  default = NULL,
+  help = "Sub count type if count_type is a list"
+)
+parser$add_argument(
+  "--sample_id_colname",
+  type = "character",
+  default = NULL,
+  help = "Column name for sample IDs"
+)
+parser$add_argument(
+  "--feature_id_colname",
+  type = "character",
+  default = NULL,
+  help = "Column name for feature IDs"
+)
+parser$add_argument(
+  "--samples_to_include",
+  type = "character",
+  default = "",
+  help = "Comma-separated list of samples to include"
+)
+parser$add_argument(
+  "--covariates_colnames",
+  type = "character",
+  default = "Group",
+  help = "Comma-separated list of covariate column names"
+)
+parser$add_argument(
+  "--contrast_colname",
+  type = "character",
+  default = "Group",
+  help = "Column containing group variables for DE"
+)
+parser$add_argument(
+  "--contrasts",
+  type = "character",
+  default = "",
+  help = "Contrasts in format group1-group2,group1-group3"
+)
+parser$add_argument(
+  "--input_in_log_counts",
+  type = "logical",
+  default = FALSE,
+  help = "Counts are already log2-transformed"
+)
+parser$add_argument(
+  "--return_mean_and_sd",
+  type = "logical",
+  default = FALSE,
+  help = "Return mean and SD in addition to DE estimates"
+)
+parser$add_argument(
+  "--voom_normalization_method",
+  type = "character",
+  default = "quantile",
+  help = "Normalization method for limma::voom"
+)
 
 args <- parser$parse_args()
 
@@ -32,25 +81,28 @@ moo <- load_moo_from_data_dir()
 
 # validate required parameters
 if (identical(args$covariates_colnames, "")) {
-    stop("covariates_colnames is required and cannot be empty")
+  stop("covariates_colnames is required and cannot be empty")
 }
 if (identical(args$contrasts, "")) {
-    stop("contrasts is required and cannot be empty")
+  stop("contrasts is required and cannot be empty")
 }
 
 # run MOSuite
-moo |> 
-    diff_counts(
-        count_type = args$count_type,
-        sub_count_type = args$sub_count_type,
-        sample_id_colname = args$sample_id_colname,
-        feature_id_colname = args$feature_id_colname,
-        samples_to_include = parse_optional_vector(args$samples_to_include),
-        covariates_colnames = parse_vector_with_default(args$covariates_colnames, "Group"),
-        contrast_colname = args$contrast_colname,
-        contrasts = parse_optional_vector(args$contrasts),
-        input_in_log_counts = args$input_in_log_counts,
-        return_mean_and_sd = args$return_mean_and_sd,
-        voom_normalization_method = args$voom_normalization_method
-        ) |> 
-    write_rds(file.path(getOption("moo_plots_dir"), "..", "moo", "moo.rds"))
+moo |>
+  diff_counts(
+    count_type = args$count_type,
+    sub_count_type = args$sub_count_type,
+    sample_id_colname = args$sample_id_colname,
+    feature_id_colname = args$feature_id_colname,
+    samples_to_include = parse_optional_vector(args$samples_to_include),
+    covariates_colnames = parse_vector_with_default(
+      args$covariates_colnames,
+      "Group"
+    ),
+    contrast_colname = args$contrast_colname,
+    contrasts = parse_optional_vector(args$contrasts),
+    input_in_log_counts = args$input_in_log_counts,
+    return_mean_and_sd = args$return_mean_and_sd,
+    voom_normalization_method = args$voom_normalization_method
+  ) |>
+  write_rds(file.path(getOption("moo_plots_dir"), "..", "moo", "moo-diff.rds"))
diff --git a/tests/data/moo-filt.rds b/tests/data/moo-filt.rds
new file mode 100644
index 0000000..24a9a52
Binary files /dev/null and b/tests/data/moo-filt.rds differ
diff --git a/tests/testthat.R b/tests/testthat.R
new file mode 100644
index 0000000..ca2904a
--- /dev/null
+++ b/tests/testthat.R
@@ -0,0 +1,2 @@
+library(testthat)
+test_dir(file.path('tests', 'testthat'), stop_on_failure = TRUE)
diff --git a/tests/testthat/test-main.R b/tests/testthat/test-main.R
new file mode 100644
index 0000000..9738a5c
--- /dev/null
+++ b/tests/testthat/test-main.R
@@ -0,0 +1,146 @@
+test_that("code/run executes successfully with default CLI arguments", {
+  # Create temporary workspace
+  workspace <- tempfile("mosuite_diff_counts_test_")
+  dir.create(workspace)
+  on.exit(unlink(workspace, recursive = TRUE), add = TRUE)
+
+  # Set up directory structure
+  code_dir <- file.path(workspace, "code")
+  data_dir <- file.path(workspace, "data")
+  results_dir <- file.path(code_dir, "..", "results")
+  dir.create(code_dir)
+  dir.create(data_dir)
+  dir.create(results_dir)
+
+  # Get test data from package tests directory
+  repo_root <- normalizePath(file.path(dirname(getwd()), ".."))
+  test_data_file <- file.path(repo_root, "tests", "data", "moo-filt.rds")
+
+  expect_true(
+    file.exists(test_data_file),
+    info = paste("Test data file should exist at", test_data_file)
+  )
+  file.copy(test_data_file, file.path(data_dir, "moo.rds"))
+
+  # Copy main.R and run script to workspace
+  file.copy(
+    file.path(repo_root, "code", "main.R"),
+    file.path(code_dir, "main.R")
+  )
+  file.copy(
+    file.path(repo_root, "code", "run"),
+    file.path(code_dir, "run")
+  )
+
+  # Run the script from code directory
+  old_wd <- getwd()
+  setwd(code_dir)
+  on.exit(setwd(old_wd), add = TRUE)
+
+  # Execute run script with default CLI arguments
+  exit_code <- system2(
+    "bash",
+    args = c(
+      "run",
+      "--covariates_colnames=Group",
+      "--contrast_colname=Group",
+      "--contrasts=B-A"
+    )
+  )
+
+  # Check for successful execution
+  expect_equal(exit_code, 0, info = "run script should execute without error")
+  expect_true(
+    file.exists(file.path(results_dir, "moo", "moo-diff.rds")),
+    info = "Output file moo-diff.rds should be created"
+  )
+
+  # Validate output is a valid MOO object
+  moo <- readr::read_rds(file.path(results_dir, "moo", "moo-diff.rds"))
+  expect_true(
+    S7::S7_inherits(moo, MOSuite::multiOmicDataSet),
+    info = "Output should be an S7 multiOmicDataSet object"
+  )
+
+  # Validate diff results exist
+  expect_true(
+    "diff" %in% names(moo@analyses),
+    info = "Output should have diff results in moo@analyses"
+  )
+})
+
+test_that("code/run executes with custom CLI arguments", {
+  # Create temporary workspace
+  workspace <- tempfile("mosuite_diff_counts_custom_test_")
+  dir.create(workspace)
+  on.exit(unlink(workspace, recursive = TRUE), add = TRUE)
+
+  # Set up directory structure
+  code_dir <- file.path(workspace, "code")
+  data_dir <- file.path(workspace, "data")
+  results_dir <- file.path(code_dir, "..", "results")
+  dir.create(code_dir)
+  dir.create(data_dir)
+  dir.create(results_dir)
+
+  # Get test data from package tests directory
+  repo_root <- normalizePath(file.path(dirname(getwd()), ".."))
+  test_data_file <- file.path(repo_root, "tests", "data", "moo-filt.rds")
+
+  # Copy test data to workspace
+  file.copy(test_data_file, file.path(data_dir, "moo.rds"))
+
+  # Copy main.R and run script to workspace
+  file.copy(
+    file.path(repo_root, "code", "main.R"),
+    file.path(code_dir, "main.R")
+  )
+  file.copy(
+    file.path(repo_root, "code", "run"),
+    file.path(code_dir, "run")
+  )
+
+  # Run the script from code directory
+  old_wd <- getwd()
+  setwd(code_dir)
+  on.exit(setwd(old_wd), add = TRUE)
+
+  # Execute run script with custom CLI arguments
+  exit_code <- system2(
+    "bash",
+    args = c(
+      "run",
+      "--count_type=filt",
+      "--covariates_colnames=Group",
+      "--contrast_colname=Group",
+      "--contrasts=C-A",
+      "--input_in_log_counts=FALSE",
+      "--return_mean_and_sd=TRUE",
+      "--voom_normalization_method=quantile"
+    )
+  )
+
+  # Check for successful execution
+  expect_equal(
+    exit_code,
+    0,
+    info = "run script with custom args should execute without error"
+  )
+  expect_true(
+    file.exists(file.path(results_dir, "moo", "moo-diff.rds")),
+    info = "Output file moo-diff.rds should be created with custom args"
+  )
+
+  # Validate output is a valid MOO object
+  moo <- readr::read_rds(file.path(results_dir, "moo", "moo-diff.rds"))
+  expect_true(
+    S7::S7_inherits(moo, MOSuite::multiOmicDataSet),
+    info = "Output should be an S7 multiOmicDataSet object"
+  )
+
+  # Validate diff results exist
+  expect_true(
+    "diff" %in% names(moo@analyses),
+    info = "Output should have diff results in moo@analyses"
+  )
+})