called "MLCI" that makes testing more convenient: it can execute any full CI job
against a pending change. Search for "MLCI" internally to find it.

You may invoke a CI script of your choice by following these instructions:

```bash
cd tensorflow-git-dir

# Here is a single-line example of running a script on Linux to build the

build:rbe_cross_compile_base_xla --config=rbe_cross_compile_base
build:rbe_cross_compile_base_xla --strategy=Genrule=standalone

# Due to the above strategy, all Genrule commands are executed locally, but the
# following actions invoke tools (E.g `flatc`, `llvm-tblgen`, etc.) that are
# only executabe on the RBE (x86) machine, so the strategy_regexp options are
# added to override and run the actions using remote strategy.

    void (*delete_func)(void*), TF_Status* status);

// If the ResourceMgr provided by the `ctx` has a resource
// `plugin_resource_name`, returns it in `*result_plugin_resource`. Otherwise,
// invokes create_func to create the resource. `delete_func` is needed for
// ResourceMgr to clean up the resource. `status` will be set. If `status` is
// not OK, `*result_plugin_resource` will be set as nullptr.
//

*   distribute.experimental.rpc package:

    *   distribute.experimental.rpc package introduces APIs to create a GRPC
        based server to register tf.function methods and a GRPC client to invoke
        remote registered methods. RPC APIs are intended for multi-client setups
        i.e. server and clients are started in separate binaries independently.

    *   Example usage to create server: ```python server =

    TF_CoordinationServiceAgent* agent, TF_Status* status) {
  if (timeout_seconds <= 0) {
    status->status = absl::InvalidArgumentError(
        "TF_CoordinationServiceGetKeyValueWithTimeout invoked with invalid "
        "timeout_seconds <= 0.");
    return nullptr;
  }
  auto* cc_agent = reinterpret_cast<tsl::CoordinationServiceAgent*>(agent);
  auto value = cc_agent->GetKeyValue(std::string_view(key, key_size),