fix: Resolve callback type mismatch in client_registry

- Fix lifetime issue in on_proc_begin callback registration
- Use cloned context in async closure to avoid lifetime conflicts
- Rename unused variable to suppress warning

1 files changed, 39 insertions, 0 deletions

diff --git a/crates/vcs_actions/src/registry/client_registry.rs b/crates/vcs_actions/src/registry/client_registry.rs
index e69de29..c32ce5a 100644
--- a/crates/vcs_actions/src/registry/client_registry.rs
+++ b/crates/vcs_actions/src/registry/client_registry.rs
@@ -0,0 +1,39 @@
+use action_system::{action::ActionContext, action_pool::ActionPool};
+use tcp_connection::error::TcpTargetError;
+
+use crate::actions::local_actions::SetUpstreamVaultAction;
+
+fn register_actions(pool: &mut ActionPool) {
+    // Pool register here
+    SetUpstreamVaultAction::register_to_pool(pool);
+}
+
+pub fn client_action_pool() -> ActionPool {
+    // Create pool
+    let mut pool = ActionPool::new();
+
+    // Register actions
+    register_actions(&mut pool);
+
+    // Add process events
+    pool.set_on_proc_begin(|ctx| Box::pin(on_proc_begin(ctx)));
+
+    // Return
+    pool
+}
+
+async fn on_proc_begin(ctx: &ActionContext) -> Result<(), TcpTargetError> {
+    // Get instance
+    let Some(_instance) = ctx.instance() else {
+        return Err(TcpTargetError::Unsupported(
+            "Missing ConnectionInstance in current context, this ActionPool does not support this call"
+                .to_string()));
+    };
+
+    // If it's remote, invoke action at server
+    if ctx.is_remote() {
+        // instance.write_text(text)
+    }
+
+    Ok(())
+}