diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index b6993e6..a543b75 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -5,8 +5,8 @@ on:
   pull_request:
 
 jobs:
-  test:
-    name: Test
+  test-action:
+    name: Test action
     strategy:
       fail-fast: false
       matrix:
@@ -291,3 +291,13 @@ jobs:
               --expect-cross-version "${{ matrix.platform.expect_cross_version }}" \
               ${{ matrix.platform.expect_cross }} \
               ${{ matrix.platform.expect_stripped }}
+
+  test-validate-inputs:
+    name: Test validate-inputs
+    runs-on: ubuntu-22.04
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - name: Run tests
+        shell: bash
+        run: ./validate-inputs.py --test
diff --git a/Changes.md b/Changes.md
index 977c851..b2f420a 100644
--- a/Changes.md
+++ b/Changes.md
@@ -6,6 +6,7 @@ bumped to v1.0.0 because of this change.
 - This action will now configure and use `Swatinem/rust-cache` by default for you. It will include
   the `target` parameter as part of the cache key automatically. Suggested by @jennydaman (Jennings
   Zhang). GH #23.
+- This action now validates its input and will exit early if they are not valid. GH #35.
 
 ## 0.0.17 - 2024-11-23
 
diff --git a/action.yml b/action.yml
index e08386d..e2d3fa9 100644
--- a/action.yml
+++ b/action.yml
@@ -6,20 +6,20 @@ branding:
 description: |
   Cross compile your Rust projects with cross (https://github.com/cross-rs/cross).
 inputs:
-  working-directory:
-    description: The working directory for each step
-    default: "."
+  target:
+    description: The target platform
+    required: true
   command:
     description: |
       The commands to run. This must be one of "build", "test", "both" (build and test), or "bench".
     default: build
-  target:
-    description: The target platform
-    required: true
   toolchain:
     description: |
       The target toolchain to use (one of "stable", "beta", or "nightly").
     default: stable
+  working-directory:
+    description: The working directory for each step
+    default: "."
   GITHUB_TOKEN:
     description: |
       A GitHub token, available in the secrets.GITHUB_TOKEN working-directory variable.
@@ -56,6 +56,19 @@ runs:
     - name: Add this action's path to PATH
       shell: bash
       run: echo "${{ github.action_path }}" >> $GITHUB_PATH
+    - name: Validate inputs
+      shell: bash
+      run: |
+        "${{ github.action_path }}"/validate-inputs.py "${{ github.workspace }}"
+      env:
+        INPUTS_target: ${{ inputs.target }}
+        INPUTS_command: ${{ inputs.command }}
+        INPUTS_toolchain: ${{ inputs.toolchain }}
+        INPUTS_working_directory: ${{ inputs.working-directory }}
+        INPUTS_strip: ${{ inputs.strip }}
+        INPUTS_cache_cross_binary: ${{ inputs.cache-cross-binary }}
+        INPUTS_use_rust_cache: ${{ inputs.use-rust-cache }}
+        INPUTS_rust_cache_parameters: ${{ inputs.rust-cache-parameters }}
     - name: Determine whether we need to cross-compile
       id: determine-cross-compile
       shell: bash
diff --git a/precious.toml b/precious.toml
index ac07b63..f8fc867 100644
--- a/precious.toml
+++ b/precious.toml
@@ -3,6 +3,24 @@ exclude = [
   "tests/lib/**/*",
 ]
 
+[commands."ruff for linting"]
+type = "both"
+include = [ "**/*.py" ]
+cmd = "ruff"
+lint_flags = [ "check" ]
+tidy_flags = [ "check", "--fix" ]
+ok_exit_codes = 0
+lint_failure_exit_codes = 1
+
+[commands."ruff for tidying"]
+type = "both"
+include = [ "**/*.py" ]
+cmd = "ruff"
+lint_flags = [ "format", "--check" ]
+tidy_flags = [ "format" ]
+ok_exit_codes = 0
+lint_failure_exit_codes = 1
+
 [commands.typos]
 type = "both"
 include = "**/*"
diff --git a/validate-inputs.py b/validate-inputs.py
new file mode 100755
index 0000000..ca955bd
--- /dev/null
+++ b/validate-inputs.py
@@ -0,0 +1,261 @@
+#!/usr/bin/env python3
+
+# Written by Claude.ai
+
+import os
+import json
+from pathlib import Path
+from typing import Dict, List, Union
+import tempfile
+import unittest
+
+
+class InputValidator:
+    """Validate inputs for a GitHub Action."""
+
+    def __init__(self, repo_root: Union[str, Path]):
+        """
+        Create a new InputValidator by collecting environment variables.
+
+        Args:
+            repo_root: Path to the repository root
+        """
+        self.repo_root = Path(repo_root)
+        self.inputs: Dict[str, str] = {
+            key.replace("INPUTS_", "").lower(): value
+            for key, value in os.environ.items()
+            if key.startswith("INPUTS_")
+        }
+
+    def validate(self) -> List[str]:
+        """
+        Validate all inputs according to specifications.
+
+        Returns:
+            List of validation errors. Empty list means all inputs are valid.
+        """
+        validation_errors: List[str] = []
+
+        # Check for required 'target' parameter
+        if "target" not in self.inputs:
+            validation_errors.append("'target' is a required parameter")
+
+        # Validate command if present
+        if "command" in self.inputs:
+            valid_commands = {"build", "test", "both", "bench"}
+            if self.inputs["command"] not in valid_commands:
+                validation_errors.append(
+                    f"Invalid 'command'. Must be one of {sorted(valid_commands)}"
+                )
+
+        # Validate toolchain if present
+        if "toolchain" in self.inputs:
+            valid_toolchains = {"stable", "beta", "nightly"}
+            if self.inputs["toolchain"] not in valid_toolchains:
+                validation_errors.append(
+                    f"Invalid 'toolchain'. Must be one of {sorted(valid_toolchains)}"
+                )
+
+        # Validate working directory if present
+        if "working_directory" in self.inputs:
+            path = Path(self.inputs["working_directory"])
+            if not path.is_absolute():
+                path = self.repo_root / path
+
+            if not path.exists():
+                validation_errors.append(
+                    f"'working-directory' does not exist: {self.inputs['working_directory']}"
+                )
+            elif not path.is_dir():
+                validation_errors.append(
+                    f"'working-directory' is not a directory: {self.inputs['working_directory']}"
+                )
+
+        # Validate boolean flags
+        boolean_flags = {"cache_cross_binary", "strip", "use_rust_cache"}
+        for flag in boolean_flags:
+            if flag in self.inputs and self.inputs[flag] not in {"true", "false"}:
+                validation_errors.append(f"'{flag}' must be either 'true' or 'false'")
+
+        # Validate rust-cache-parameters JSON if present
+        if "rust_cache_parameters" in self.inputs:
+            try:
+                json.loads(self.inputs["rust_cache_parameters"])
+            except json.JSONDecodeError:
+                validation_errors.append("'rust-cache-parameters' must be valid JSON")
+
+        return validation_errors
+
+
+def main() -> None:
+    """Main function for running the validator."""
+    import sys
+
+    validator = InputValidator(sys.argv[1])
+    errors = validator.validate()
+
+    if not errors:
+        print("All inputs are valid.")
+        sys.exit(0)
+    else:
+        for error in errors:
+            print(error, file=sys.stderr)
+        sys.exit(1)
+
+
+class TestInputValidator(unittest.TestCase):
+    """Unit tests for the InputValidator."""
+
+    def setUp(self) -> None:
+        """Set up test environment."""
+        # Clear existing INPUTS_ environment variables
+        for key in list(os.environ.keys()):
+            if key.startswith("INPUTS_"):
+                del os.environ[key]
+
+    def setup_env(self, inputs: Dict[str, str]) -> None:
+        """Helper function to set up environment variables for testing."""
+        for key, value in inputs.items():
+            env_key = f"INPUTS_{key.upper().replace('-', '_')}"
+            os.environ[env_key] = value
+
+    def test_get_inputs_from_env(self) -> None:
+        """Test getting inputs from environment variables."""
+        inputs = {
+            "target": "x86_64-unknown-linux-gnu",
+            "command": "build",
+            "toolchain": "stable",
+            "use-rust-cache": "true",
+        }
+        self.setup_env(inputs)
+
+        validator = InputValidator("/root")
+        for key, value in validator.inputs.items():
+            self.assertEqual(value, inputs[key.replace("_", "-")])
+
+    def test_validate_missing_target(self) -> None:
+        """Test validation with missing target."""
+        self.setup_env({})
+        validator = InputValidator("/root")
+        errors = validator.validate()
+        self.assertTrue(errors)
+
+    def test_validate_valid_command(self) -> None:
+        """Test validation of valid commands."""
+        valid_commands = ["build", "test", "both", "bench"]
+
+        for command in valid_commands:
+            self.setup_env({"target": "x86_64-unknown-linux-gnu", "command": command})
+            validator = InputValidator("/root")
+            errors = validator.validate()
+            self.assertFalse(errors, f"Command '{command}' should be valid")
+
+    def test_validate_invalid_command(self) -> None:
+        """Test validation of invalid command."""
+        self.setup_env({"target": "x86_64-unknown-linux-gnu", "command": "invalid"})
+        validator = InputValidator("/root")
+        errors = validator.validate()
+        self.assertTrue(errors)
+
+    def test_validate_valid_toolchain(self) -> None:
+        """Test validation of valid toolchains."""
+        valid_toolchains = ["stable", "beta", "nightly"]
+
+        for toolchain in valid_toolchains:
+            self.setup_env(
+                {"target": "x86_64-unknown-linux-gnu", "toolchain": toolchain}
+            )
+            validator = InputValidator("/root")
+            errors = validator.validate()
+            self.assertFalse(errors, f"Toolchain '{toolchain}' should be valid")
+
+    def test_validate_invalid_toolchain(self) -> None:
+        """Test validation of invalid toolchain."""
+        self.setup_env({"target": "x86_64-unknown-linux-gnu", "toolchain": "unknown"})
+        validator = InputValidator("/root")
+        errors = validator.validate()
+        self.assertTrue(errors)
+
+    def test_validate_working_directory(self) -> None:
+        """Test validation of working directory."""
+        with tempfile.TemporaryDirectory() as temp_dir:
+            # Test with valid directory
+            self.setup_env(
+                {"target": "x86_64-unknown-linux-gnu", "working-directory": temp_dir}
+            )
+            validator = InputValidator("/root")
+            errors = validator.validate()
+            self.assertFalse(errors)
+
+            # Test with non-existent directory
+            self.setup_env(
+                {
+                    "target": "x86_64-unknown-linux-gnu",
+                    "working-directory": "/path/to/nonexistent/directory",
+                }
+            )
+            validator = InputValidator("/root")
+            errors = validator.validate()
+            self.assertTrue(errors)
+
+            # Test with file instead of directory
+            with tempfile.NamedTemporaryFile() as temp_file:
+                self.setup_env(
+                    {
+                        "target": "x86_64-unknown-linux-gnu",
+                        "working-directory": temp_file.name,
+                    }
+                )
+                validator = InputValidator("/root")
+                errors = validator.validate()
+                self.assertTrue(errors)
+
+    def test_validate_boolean_flags(self) -> None:
+        """Test validation of boolean flags."""
+        boolean_flags = ["cache-cross-binary", "strip", "use-rust-cache"]
+
+        # Test valid boolean values
+        for flag in boolean_flags:
+            for value in ["true", "false"]:
+                self.setup_env({"target": "x86_64-unknown-linux-gnu", flag: value})
+                validator = InputValidator("/root")
+                errors = validator.validate()
+                self.assertFalse(errors, f"'{flag}' with '{value}' should be valid")
+
+        # Test invalid boolean values
+        for flag in boolean_flags:
+            self.setup_env({"target": "x86_64-unknown-linux-gnu", flag: "invalid"})
+            validator = InputValidator("/root")
+            errors = validator.validate()
+            self.assertTrue(errors, f"'{flag}' with 'invalid' should be invalid")
+
+    def test_validate_rust_cache_parameters(self) -> None:
+        """Test validation of rust cache parameters."""
+        # Valid JSON
+        self.setup_env(
+            {
+                "target": "x86_64-unknown-linux-gnu",
+                "rust-cache-parameters": '{"key1":"value1","key2":"value2"}',
+            }
+        )
+        validator = InputValidator("/root")
+        errors = validator.validate()
+        self.assertFalse(errors)
+
+        # Invalid JSON
+        self.setup_env(
+            {
+                "target": "x86_64-unknown-linux-gnu",
+                "rust-cache-parameters": "{invalid json",
+            }
+        )
+        validator = InputValidator("/root")
+        errors = validator.validate()
+        self.assertTrue(errors)
+
+
+if __name__ == "__main__":
+    if len(os.sys.argv) > 1 and os.sys.argv[1] == "--test":
+        unittest.main(argv=["unittest"])
+    else:
+        main()