Merge pull request #1 from eval-protocol/codex/migrate-reward-kit-to-eval-protocol

Fix naming to Eval Protocol

Author: benjibc

.github/ISSUE_TEMPLATE/bug_report.md

@@ -25,9 +25,9 @@ If applicable, add screenshots to help explain your problem.
 
 **Environment (please complete the following information):**
  - OS: [e.g. macOS, Windows, Linux]
- - Python version: [e.g. 3.9, 3.10]
- - Reward Kit version: [e.g. 0.1.0, or commit SHA if from source]
- - How installed: [e.g. pip, from source]
+- Python version: [e.g. 3.9, 3.10]
+ - Eval Protocol version: [e.g. 0.1.0, or commit SHA if from source]
+- How installed: [e.g. pip, from source]
 
 **Additional context**
 Add any other context about the problem here. For example, are you using it with a specific LLM provider, or in a particular environment?

development/CONTRIBUTING.md

@@ -360,7 +360,7 @@ uv run black eval_protocol
 
 ## Available Reward Functions
 
-Reward Kit includes these out-of-the-box reward functions:
+Eval Protocol includes these out-of-the-box reward functions:
 
 | Category | Reward Functions |
 |----------|-----------------|
@@ -380,7 +380,7 @@ Reward Kit includes these out-of-the-box reward functions:
 
 ## Running Examples
 
-The examples folder contains sample code for using the Reward Kit:
+The examples folder contains sample code for using Eval Protocol:
 
 ```bash
 # Run evaluation preview example
@@ -506,7 +506,7 @@ If you encounter authentication issues:
 3.  **Check Account ID**: Confirm that the `FIREWORKS_ACCOUNT_ID` is correct for the environment you are targeting (e.g., `pyroworks-dev` for the dev API, or your personal account ID).
 4.  **API Base URL**: If using `FIREWORKS_API_BASE`, ensure it points to the correct API endpoint (e.g., `https://dev.api.fireworks.ai` for development).
 
-You can use the following snippet to check what credentials the Reward Kit is resolving:
+You can use the following snippet to check what credentials Eval Protocol is resolving:
 ```python
 from eval_protocol.auth import get_fireworks_api_key, get_fireworks_account_id
 
@@ -556,7 +556,7 @@ uv pip install dist/eval_protocol-*.whl
 
 ## Contributing Process
 
-We welcome contributions to Reward Kit! Please follow these steps to contribute:
+We welcome contributions to Eval Protocol! Please follow these steps to contribute:
 
 1.  **Find or Create an Issue**:
     *   Look for existing issues on the [GitHub Issues page](https://github.com/fireworks-ai/eval-protocol/issues) that you'd like to work on.
@@ -581,9 +581,9 @@ We welcome contributions to Reward Kit! Please follow these steps to contribute:
 
 6.  **Run Code Quality Checks**:
 
-    *   Format your code: `uv run black reward_kit tests`
-    *   Check linting: `uv run flake8 reward_kit tests`
-    *   Check types: `uv run mypy reward_kit`
+    *   Format your code: `uv run black eval_protocol tests`
+    *   Check linting: `uv run flake8 eval_protocol tests`
+    *   Check types: `uv run mypy eval_protocol`
     *   Run pre-commit hooks: `pre-commit run --all-files`
 
 7.  **Update Documentation**:

eval_protocol/__main__.py

@@ -1,5 +1,5 @@
 """
-Main entry point for running reward-kit as a module.
+Main entry point for running Eval Protocol as a module.
 """
 
 import sys

eval_protocol/_version.py

@@ -51,7 +51,7 @@ def get_config() -> VersioneerConfig:
     cfg.VCS = "git"
     cfg.style = "pep440"
     cfg.tag_prefix = "v"
-    cfg.parentdir_prefix = "reward-kit-"
+    cfg.parentdir_prefix = "eval-protocol-"
     cfg.versionfile_source = "eval_protocol/_version.py"
     cfg.verbose = False
     return cfg

eval_protocol/cli_commands/deploy.py

@@ -93,7 +93,7 @@ def _establish_local_server_and_tunnel(args):
     function_ref = args.function_ref
     local_server_port = args.local_port
 
-    log_dir = os.path.join(os.getcwd(), "logs", "reward-kit-local")
+    log_dir = os.path.join(os.getcwd(), "logs", "eval-protocol-local")
     os.makedirs(log_dir, exist_ok=True)
     generic_server_log_path = os.path.join(log_dir, f"generic_server_{evaluator_id}.log")
 
@@ -270,9 +270,9 @@ def _deploy_to_gcp_cloud_run(args, current_config, gcp_config_from_yaml):
 
         gcp_sanitized_eval_id = "".join(filter(lambda char: char.isalnum() or char in ["-", "_"], args.id))
         if not gcp_sanitized_eval_id:
-            gcp_sanitized_eval_id = "rewardkit-evaluator"
+            gcp_sanitized_eval_id = "evalprotocol-evaluator"
         secret_id_for_auth_key = f"rk-eval-{gcp_sanitized_eval_id}-authkey"
-        secret_labels = {"managed-by": "reward-kit", "evaluator-id": evaluator_id}
+        secret_labels = {"managed-by": "eval-protocol", "evaluator-id": evaluator_id}
 
         api_key_secret_version_id = ensure_gcp_secret(
             project_id=gcp_project_id,

eval_protocol/cli_commands/deploy_mcp.py

@@ -43,7 +43,7 @@ def _generate_mcp_dockerfile_content(
     # Base requirements for MCP servers - matching setup.py dependencies
     base_requirements = [
         "fastmcp>=0.1.0",
-        # Core reward-kit dependencies from setup.py
+        # Core Eval Protocol dependencies from setup.py
         "requests>=2.25.0",
         "pydantic>=2.0.0",
         "dataclasses-json>=0.5.7",
@@ -154,7 +154,7 @@ def _deploy_mcp_to_gcp_cloud_run(args, current_config, gcp_config_from_yaml):
     if not gcp_ar_repo_name and gcp_config_from_yaml:
         gcp_ar_repo_name = gcp_config_from_yaml.artifact_registry_repository
     if not gcp_ar_repo_name:
-        gcp_ar_repo_name = "reward-kit-mcp-servers"
+        gcp_ar_repo_name = "eval-protocol-mcp-servers"
 
     print(f"Using GCP Project: {gcp_project_id}, Region: {gcp_region}, AR Repo: {gcp_ar_repo_name}")
 

eval_protocol/execution/pipeline.py

@@ -1,5 +1,5 @@
 """
-Core evaluation execution pipeline for reward-kit.
+Core evaluation execution pipeline for Eval Protocol.
 This module orchestrates dataset loading, model response generation (optional),
 and evaluation using specified reward functions.
 """

eval_protocol/gcp_tools.py

@@ -289,7 +289,7 @@ def ensure_artifact_registry_repo_exists(project_id: str, region: str, repo_name
                 "--location",
                 region,
                 "--description",
-                "Repository for reward-kit evaluators (auto-created by reward-kit CLI)",
+                "Repository for Eval Protocol evaluators (auto-created by Eval Protocol CLI)",
             ]
             create_success, create_stdout, create_stderr = _run_gcloud_command(create_cmd, dry_run=dry_run)
             if create_success:
@@ -442,7 +442,7 @@ def ensure_gcp_secret(
     dummy_dockerfile = 'FROM python:3.10-slim\nCMD ["echo", "hello"]'
     img_name = "gcr.io/my-test-project/my-test-reward-eval:latest"  # Old GCR name, update for AR
     # Example AR image name: us-central1-docker.pkg.dev/my-test-project/my-ar-repo/my-test-reward-eval:latest
-    ar_img_name = "us-central1-docker.pkg.dev/my-test-project/reward-kit-images/my-test-reward-eval:latest"
+    ar_img_name = "us-central1-docker.pkg.dev/my-test-project/eval-protocol-images/my-test-reward-eval:latest"
 
     print(f"\n1. Simulating build and push for {ar_img_name} (dry_run=True)")
     build_and_push_docker_image(
@@ -478,7 +478,7 @@ def ensure_gcp_secret(
         project_id="my-test-project",
         secret_id="my-test-api-key-secret",
         secret_value="supersecretvalue123",
-        labels={"managed-by": "reward-kit-test"},
+        labels={"managed-by": "eval-protocol-test"},
         dry_run=True,
     )
     print("\nNote: These are placeholder executions. Real implementation requires gcloud CLI and Docker.")

eval_protocol/mcp_agent/config.py

@@ -114,7 +114,7 @@ def model_post_init(self, __context: Any) -> None:
 
 
 class AppConfig(BaseModel):
-    """Root configuration for the RewardKit Intermediary MCP Server."""
+    """Root configuration for the Eval Protocol Intermediary MCP Server."""
 
     backends: List[BackendServerConfig] = Field(
         default_factory=list,

eval_protocol/mcp_env.py

@@ -1,5 +1,5 @@
 """
-MCP Environment API for reward-kit - Backward Compatibility Facade
+MCP Environment API for Eval Protocol - Backward Compatibility Facade
 
 This module has been refactored into modular components for better maintainability.
 This file now serves as a backward compatibility facade.