fix(deployment): make reconciliation leases durable (#205)

## Summary

- persist delayed reconciliation wakeups instead of repeatedly
reacquiring work that is not ready
- keep eligible controller work under one bounded lease and run
independent ready resources concurrently
- renew ownership through a dedicated endpoint without rewriting
deployment state
- cancel promptly when ownership is lost and clear scheduled delays for
explicit user actions

## Why

A fresh multi-resource deployment exposed two framework failures: short
controller waits were repeatedly yielded through the shared scheduler,
and manager replicas could remain occupied retrying work after lease
loss. This preserves durable checkpoints while avoiding unnecessary
distributed scheduling between adjacent controller steps.

## Validation

- `cargo check -p alien-infra -p alien-deployment -p alien-manager
--all-features`
- 851 tests across `alien-infra`, `alien-deployment`, and
`alien-manager` passed (1 skipped)
- generated OpenAPI artifacts were produced through the repository
generators

## Dependency

Stacked on #179.

Linear:
https://linear.app/alienplatform/issue/ALIEN-324/make-deployment-leases-cancellable-and-keep-short-controller

Author: alongubkin

client-sdks/manager/openapi.json

@@ -1464,7 +1464,7 @@ pub async fn deploy_task(args: DeployArgs, ctx: ExecutionMode) -> Result<()> {
         } else {
             LoopOperation::Deploy
         },
-        delay_threshold: None,
+        delay_strategy: alien_deployment::runner::DelayStrategy::Inline,
     };
 
     let runner_result = alien_deployment::runner::run_step_loop(

client-sdks/manager/rust/openapi-3.0.json

@@ -281,7 +281,7 @@ pub async fn destroy_task(args: DestroyArgs, ctx: ExecutionMode) -> Result<()> {
     let policy = RunnerPolicy {
         max_steps: 400,
         operation: LoopOperation::Delete,
-        delay_threshold: None,
+        delay_strategy: alien_deployment::runner::DelayStrategy::Inline,
     };
 
     let runner_result = match alien_deployment::runner::run_step_loop(

client-sdks/platform/openapi.json

@@ -3576,7 +3576,7 @@ pub async fn push_initial_setup(
         // The CLI drives Pending → InitialSetup → Provisioning, then stops.
         // The manager picks up from Provisioning and drives to Running.
         operation: LoopOperation::InitialSetup,
-        delay_threshold: None,
+        delay_strategy: alien_deployment::runner::DelayStrategy::Inline,
     };
 
     let runner_result = shared_run_step_loop(
@@ -3840,7 +3840,7 @@ async fn run_runtime_deletion(
     let policy = RunnerPolicy {
         max_steps: 400,
         operation: LoopOperation::Delete,
-        delay_threshold: None,
+        delay_strategy: alien_deployment::runner::DelayStrategy::Inline,
     };
 
     let runner_result = shared_run_step_loop(
@@ -3938,7 +3938,7 @@ async fn run_setup_deletion(
     let policy = RunnerPolicy {
         max_steps: 400,
         operation: LoopOperation::Delete,
-        delay_threshold: None,
+        delay_strategy: alien_deployment::runner::DelayStrategy::Inline,
     };
 
     let runner_result = alien_deployment::setup_teardown::run_setup_teardown_after_handoff(

client-sdks/platform/rust/openapi-3.0.json

@@ -22,6 +22,8 @@ pub enum LoopStopReason {
     Handoff,
     /// No work was available (no target release, no config, etc.).
     NoWork,
+    /// The next useful reconciliation is scheduled for later.
+    Delayed,
     /// Step budget exhausted without reaching a terminal state.
     BudgetExceeded,
     /// Loop was cancelled by an external signal.

client-sdks/platform/rust/openapi.json

@@ -38,6 +38,22 @@ impl ManagerApiTransport {
 
 #[async_trait]
 impl DeploymentLoopTransport for ManagerApiTransport {
+    async fn renew_lease(&self, deployment_id: &str) -> Result<(), AlienError> {
+        self.client
+            .renew()
+            .body(alien_manager_api::types::RenewRequest {
+                deployment_id: deployment_id.to_string(),
+                session: self.session.clone(),
+            })
+            .send()
+            .await
+            .into_sdk_error()
+            .context(alien_error::GenericError {
+                message: "Failed to renew deployment lease via manager API".to_string(),
+            })?;
+        Ok(())
+    }
+
     async fn reconcile_step(
         &self,
         deployment_id: &str,