Allow setting custom inference timeout

architecture/inference-routing.md

@@ -92,10 +92,10 @@ File: `proto/inference.proto`
 
 Key messages:
 
-- `SetClusterInferenceRequest` -- `provider_name` + `model_id` + optional `no_verify` override, with verification enabled by default
-- `SetClusterInferenceResponse` -- `provider_name` + `model_id` + `version`
+- `SetClusterInferenceRequest` -- `provider_name` + `model_id` + `timeout_secs` + optional `no_verify` override, with verification enabled by default
+- `SetClusterInferenceResponse` -- `provider_name` + `model_id` + `timeout_secs` + `version`
 - `GetInferenceBundleResponse` -- `repeated ResolvedRoute routes` + `revision` + `generated_at_ms`
-- `ResolvedRoute` -- `name`, `base_url`, `protocols`, `api_key`, `model_id`, `provider_type`
+- `ResolvedRoute` -- `name`, `base_url`, `protocols`, `api_key`, `model_id`, `provider_type`, `timeout_secs`
 
 ## Data Plane (Sandbox)
 
@@ -106,7 +106,7 @@ Files:
 - `crates/openshell-sandbox/src/lib.rs` -- inference context initialization, route refresh
 - `crates/openshell-sandbox/src/grpc_client.rs` -- `fetch_inference_bundle()`
 
-In cluster mode, the sandbox starts a background refresh loop as soon as the inference context is created. The loop polls the gateway every 5 seconds by default (`OPENSHELL_ROUTE_REFRESH_INTERVAL_SECS` override) and uses the bundle revision hash to skip no-op cache writes.
+In cluster mode, the sandbox starts a background refresh loop as soon as the inference context is created. The loop polls the gateway every 5 seconds by default (`OPENSHELL_ROUTE_REFRESH_INTERVAL_SECS` override) and uses the bundle revision hash to skip no-op cache writes. The revision hash covers all route fields including `timeout_secs`, so any configuration change (provider, model, or timeout) triggers a cache update on the next poll.
 
 ### Interception flow
 
@@ -143,7 +143,7 @@ If no pattern matches, the proxy returns `403 Forbidden` with `{"error": "connec
 ### Route cache
 
 - `InferenceContext` holds a `Router`, the pattern list, and an `Arc<RwLock<Vec<ResolvedRoute>>>` route cache.
-- In cluster mode, `spawn_route_refresh()` polls `GetInferenceBundle` every 30 seconds (`ROUTE_REFRESH_INTERVAL_SECS`). On failure, stale routes are kept.
+- In cluster mode, `spawn_route_refresh()` polls `GetInferenceBundle` every 5 seconds (`OPENSHELL_ROUTE_REFRESH_INTERVAL_SECS`). On failure, stale routes are kept.
 - In file mode (`--inference-routes`), routes load once at startup from YAML. No refresh task is spawned.
 - In cluster mode, an empty initial bundle still enables the inference context so the refresh task can pick up later configuration.
 
@@ -209,9 +209,11 @@ File: `crates/openshell-router/src/mock.rs`
 
 Routes with `mock://` scheme endpoints return canned responses without making HTTP requests. Mock responses are protocol-aware (OpenAI chat completion, OpenAI completion, Anthropic messages, or generic JSON). Mock routes include an `x-openshell-mock: true` response header.
 
-### HTTP client
+### Per-request timeout
 
-The router uses a `reqwest::Client` with a 60-second timeout. Timeouts and connection failures map to `RouterError::UpstreamUnavailable`.
+Each `ResolvedRoute` carries a `timeout` field (`Duration`). The `reqwest::Client` has no global timeout; instead, each outgoing request applies `.timeout(route.timeout)` on the request builder. When `timeout_secs` is `0` in the proto message, the default of 60 seconds is used (defined as `DEFAULT_ROUTE_TIMEOUT` in `config.rs`). Timeouts and connection failures map to `RouterError::UpstreamUnavailable`.
+
+Timeout changes propagate dynamically to running sandboxes. The bundle revision hash includes `timeout_secs`, so when the timeout is updated via `openshell inference update --timeout`, the refresh loop detects the revision change and updates the route cache within one polling interval (5 seconds by default).
 
 ## Standalone Route File
 
@@ -297,13 +299,16 @@ The system route is stored as a separate `InferenceRoute` record in the gateway
 
 Cluster inference commands:
 
-- `openshell inference set --provider <name> --model <id>` -- configures user-facing cluster inference
-- `openshell inference set --system --provider <name> --model <id>` -- configures system inference
+- `openshell inference set --provider <name> --model <id> [--timeout <secs>]` -- configures user-facing cluster inference
+- `openshell inference set --system --provider <name> --model <id> [--timeout <secs>]` -- configures system inference
+- `openshell inference update [--provider <name>] [--model <id>] [--timeout <secs>]` -- updates individual fields without resetting others
 - `openshell inference get` -- displays both user and system inference configuration
 - `openshell inference get --system` -- displays only the system inference configuration
 
 The `--provider` flag references a provider record name (not a provider type). The provider must already exist in the cluster and have a supported inference type (`openai`, `anthropic`, or `nvidia`).
 
+The `--timeout` flag sets the per-request timeout in seconds for upstream inference calls. When omitted or set to `0`, the default of 60 seconds applies. Timeout changes propagate to running sandboxes within the route refresh interval (5 seconds by default).
+
 Inference writes verify by default. `--no-verify` is the explicit opt-out for endpoints that are not up yet.
 
 ## Provider Discovery

crates/openshell-cli/src/main.rs

@@ -937,6 +937,10 @@ enum InferenceCommands {
         /// Skip endpoint verification before saving the route.
         #[arg(long)]
         no_verify: bool,
+
+        /// Request timeout in seconds for inference calls (0 = default 60s).
+        #[arg(long, default_value_t = 0)]
+        timeout: u64,
     },
 
     /// Update gateway-level inference configuration (partial update).
@@ -957,6 +961,10 @@ enum InferenceCommands {
         /// Skip endpoint verification before saving the route.
         #[arg(long)]
         no_verify: bool,
+
+        /// Request timeout in seconds for inference calls (0 = default 60s, unchanged if omitted).
+        #[arg(long)]
+        timeout: Option<u64>,
     },
 
     /// Get gateway-level inference provider and model.
@@ -2026,10 +2034,12 @@ async fn main() -> Result<()> {
                     model,
                     system,
                     no_verify,
+                    timeout,
                 } => {
                     let route_name = if system { "sandbox-system" } else { "" };
                     run::gateway_inference_set(
-                        endpoint, &provider, &model, route_name, no_verify, &tls,
+                        endpoint, &provider, &model, route_name, no_verify, timeout,
+                        &tls,
                     )
                     .await?;
                 }
@@ -2038,6 +2048,7 @@ async fn main() -> Result<()> {
                     model,
                     system,
                     no_verify,
+                    timeout,
                 } => {
                     let route_name = if system { "sandbox-system" } else { "" };
                     run::gateway_inference_update(
@@ -2046,6 +2057,7 @@ async fn main() -> Result<()> {
                         model.as_deref(),
                         route_name,
                         no_verify,
+                        timeout,
                         &tls,
                     )
                     .await?;

crates/openshell-cli/src/run.rs

@@ -3481,6 +3481,7 @@ pub async fn gateway_inference_set(
     model_id: &str,
     route_name: &str,
     no_verify: bool,
+    timeout_secs: u64,
     tls: &TlsOptions,
 ) -> Result<()> {
     let progress = if std::io::stdout().is_terminal() {
@@ -3504,6 +3505,7 @@ pub async fn gateway_inference_set(
             route_name: route_name.to_string(),
             verify: false,
             no_verify,
+            timeout_secs,
         })
         .await;
 
@@ -3525,6 +3527,7 @@ pub async fn gateway_inference_set(
     println!("  {} {}", "Provider:".dimmed(), configured.provider_name);
     println!("  {} {}", "Model:".dimmed(), configured.model_id);
     println!("  {} {}", "Version:".dimmed(), configured.version);
+    print_timeout(configured.timeout_secs);
     if configured.validation_performed {
         println!("  {}", "Validated Endpoints:".dimmed());
         for endpoint in configured.validated_endpoints {
@@ -3540,11 +3543,12 @@ pub async fn gateway_inference_update(
     model_id: Option<&str>,
     route_name: &str,
     no_verify: bool,
+    timeout_secs: Option<u64>,
     tls: &TlsOptions,
 ) -> Result<()> {
-    if provider_name.is_none() && model_id.is_none() {
+    if provider_name.is_none() && model_id.is_none() && timeout_secs.is_none() {
         return Err(miette::miette!(
-            "at least one of --provider or --model must be specified"
+            "at least one of --provider, --model, or --timeout must be specified"
         ));
     }
 
@@ -3561,6 +3565,7 @@ pub async fn gateway_inference_update(
 
     let provider = provider_name.unwrap_or(&current.provider_name);
     let model = model_id.unwrap_or(&current.model_id);
+    let timeout = timeout_secs.unwrap_or(current.timeout_secs);
 
     let progress = if std::io::stdout().is_terminal() {
         let spinner = ProgressBar::new_spinner();
@@ -3582,6 +3587,7 @@ pub async fn gateway_inference_update(
             route_name: route_name.to_string(),
             verify: false,
             no_verify,
+            timeout_secs: timeout,
         })
         .await;
 
@@ -3603,6 +3609,7 @@ pub async fn gateway_inference_update(
     println!("  {} {}", "Provider:".dimmed(), configured.provider_name);
     println!("  {} {}", "Model:".dimmed(), configured.model_id);
     println!("  {} {}", "Version:".dimmed(), configured.version);
+    print_timeout(configured.timeout_secs);
     if configured.validation_performed {
         println!("  {}", "Validated Endpoints:".dimmed());
         for endpoint in configured.validated_endpoints {
@@ -3639,6 +3646,7 @@ pub async fn gateway_inference_get(
         println!("  {} {}", "Provider:".dimmed(), configured.provider_name);
         println!("  {} {}", "Model:".dimmed(), configured.model_id);
         println!("  {} {}", "Version:".dimmed(), configured.version);
+        print_timeout(configured.timeout_secs);
     } else {
         // Show both routes by default.
         print_inference_route(&mut client, "Gateway inference", "").await;
@@ -3666,6 +3674,7 @@ async fn print_inference_route(
             println!("  {} {}", "Provider:".dimmed(), configured.provider_name);
             println!("  {} {}", "Model:".dimmed(), configured.model_id);
             println!("  {} {}", "Version:".dimmed(), configured.version);
+            print_timeout(configured.timeout_secs);
         }
         Err(e) if e.code() == Code::NotFound => {
             println!("{}", format!("{label}:").cyan().bold());
@@ -3680,6 +3689,14 @@ async fn print_inference_route(
     }
 }
 
+fn print_timeout(timeout_secs: u64) {
+    if timeout_secs == 0 {
+        println!("  {} {}s (default)", "Timeout:".dimmed(), 60);
+    } else {
+        println!("  {} {}s", "Timeout:".dimmed(), timeout_secs);
+    }
+}
+
 fn format_inference_status(status: Status) -> miette::Report {
     let message = status.message().trim();
 

crates/openshell-router/src/backend.rs

@@ -149,7 +149,7 @@ async fn send_backend_request(
         }
         Err(_) => body,
     };
-    builder = builder.body(body);
+    builder = builder.body(body).timeout(route.timeout);
 
     builder.send().await.map_err(|e| {
         if e.is_timeout() {
@@ -468,6 +468,7 @@ mod tests {
             protocols: protocols.iter().map(|p| (*p).to_string()).collect(),
             auth,
             default_headers: vec![("anthropic-version".to_string(), "2023-06-01".to_string())],
+            timeout: crate::config::DEFAULT_ROUTE_TIMEOUT,
         }
     }
 

crates/openshell-router/src/config.rs

@@ -3,11 +3,14 @@
 
 use serde::Deserialize;
 use std::path::Path;
+use std::time::Duration;
 
 pub use openshell_core::inference::AuthHeader;
 
 use crate::RouterError;
 
+pub const DEFAULT_ROUTE_TIMEOUT: Duration = Duration::from_secs(60);
+
 #[derive(Debug, Clone, Deserialize)]
 pub struct RouterConfig {
     pub routes: Vec<RouteConfig>,
@@ -45,6 +48,8 @@ pub struct ResolvedRoute {
     pub auth: AuthHeader,
     /// Extra headers injected on every request (e.g. `anthropic-version`).
     pub default_headers: Vec<(String, String)>,
+    /// Per-request timeout for proxied inference calls.
+    pub timeout: Duration,
 }
 
 impl std::fmt::Debug for ResolvedRoute {
@@ -57,6 +62,7 @@ impl std::fmt::Debug for ResolvedRoute {
             .field("protocols", &self.protocols)
             .field("auth", &self.auth)
             .field("default_headers", &self.default_headers)
+            .field("timeout", &self.timeout)
             .finish()
     }
 }
@@ -129,6 +135,7 @@ impl RouteConfig {
             protocols,
             auth,
             default_headers,
+            timeout: DEFAULT_ROUTE_TIMEOUT,
         })
     }
 }
@@ -256,6 +263,7 @@ routes:
             protocols: vec!["openai_chat_completions".to_string()],
             auth: AuthHeader::Bearer,
             default_headers: Vec::new(),
+            timeout: DEFAULT_ROUTE_TIMEOUT,
         };
         let debug_output = format!("{route:?}");
         assert!(

crates/openshell-router/src/lib.rs

@@ -5,8 +5,6 @@ mod backend;
 pub mod config;
 mod mock;
 
-use std::time::Duration;
-
 pub use backend::{
     ProxyResponse, StreamingProxyResponse, ValidatedEndpoint, ValidationFailure,
     ValidationFailureKind, verify_backend_endpoint,
@@ -39,7 +37,6 @@ pub struct Router {
 impl Router {
     pub fn new() -> Result<Self, RouterError> {
         let client = reqwest::Client::builder()
-            .timeout(Duration::from_secs(60))
             .build()
             .map_err(|e| RouterError::Internal(format!("failed to build HTTP client: {e}")))?;
         Ok(Self {

crates/openshell-router/src/mock.rs

@@ -131,6 +131,7 @@ mod tests {
             protocols: protocols.iter().map(ToString::to_string).collect(),
             auth: crate::config::AuthHeader::Bearer,
             default_headers: Vec::new(),
+            timeout: crate::config::DEFAULT_ROUTE_TIMEOUT,
         }
     }
 

crates/openshell-router/tests/backend_integration.rs

@@ -15,6 +15,7 @@ fn mock_candidates(base_url: &str) -> Vec<ResolvedRoute> {
         protocols: vec!["openai_chat_completions".to_string()],
         auth: AuthHeader::Bearer,
         default_headers: Vec::new(),
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }]
 }
 
@@ -117,6 +118,7 @@ async fn proxy_no_compatible_route_returns_error() {
         protocols: vec!["anthropic_messages".to_string()],
         auth: AuthHeader::Custom("x-api-key"),
         default_headers: Vec::new(),
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }];
 
     let err = router
@@ -178,6 +180,7 @@ async fn proxy_mock_route_returns_canned_response() {
         protocols: vec!["openai_chat_completions".to_string()],
         auth: AuthHeader::Bearer,
         default_headers: Vec::new(),
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }];
 
     let body = serde_json::to_vec(&serde_json::json!({
@@ -312,6 +315,7 @@ async fn proxy_uses_x_api_key_for_anthropic_route() {
         protocols: vec!["anthropic_messages".to_string()],
         auth: AuthHeader::Custom("x-api-key"),
         default_headers: vec![("anthropic-version".to_string(), "2023-06-01".to_string())],
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }];
 
     let body = serde_json::to_vec(&serde_json::json!({
@@ -370,6 +374,7 @@ async fn proxy_anthropic_does_not_send_bearer_auth() {
         protocols: vec!["anthropic_messages".to_string()],
         auth: AuthHeader::Custom("x-api-key"),
         default_headers: vec![("anthropic-version".to_string(), "2023-06-01".to_string())],
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }];
 
     let response = router
@@ -414,6 +419,7 @@ async fn proxy_forwards_client_anthropic_version_header() {
         protocols: vec!["anthropic_messages".to_string()],
         auth: AuthHeader::Custom("x-api-key"),
         default_headers: vec![("anthropic-version".to_string(), "2023-06-01".to_string())],
+        timeout: openshell_router::config::DEFAULT_ROUTE_TIMEOUT,
     }];
 
     let body = serde_json::to_vec(&serde_json::json!({