Add 'metrics test' subcommand for testing metrics against simulation outputs

src/agent_discovery.rs

@@ -412,7 +412,7 @@ const RESOURCE_SPECS: &[ResourceSpec] = &[
     },
     ResourceSpec {
         name: "metrics",
-        commands: &["context", "list", "get", "create", "update", "delete"],
+        commands: &["context", "list", "get", "create", "update", "delete", "test"],
         description: "Metrics score conversations and simulations with built-in or configured evaluation logic.",
         id_name: "metric_id",
         id_format: "22-character Coval ID",

src/client/mod.rs

@@ -628,6 +628,15 @@ impl MetricsClient<'_> {
         let url = self.0.url(&format!("/v1/metrics/{id}"));
         self.0.delete(url).await
     }
+
+    pub async fn test(
+        &self,
+        id: &str,
+        req: models::TestMetricRequest,
+    ) -> Result<models::TestMetricResponse, ApiError> {
+        let url = self.0.url(&format!("/v1/metrics/{id}/test"));
+        self.0.post(url, &req).await
+    }
 }
 
 impl MutationsClient<'_> {

src/client/models/metric.rs

@@ -219,6 +219,18 @@ pub struct UpdateMetricResponse {
     pub metric: Metric,
 }
 
+#[derive(Debug, Serialize)]
+pub struct TestMetricRequest {
+    pub simulation_output_id: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub dev_id: Option<String>,
+}
+
+#[derive(Debug, Deserialize, Serialize)]
+pub struct TestMetricResponse {
+    pub metric_output_ulid: String,
+}
+
 impl Tabular for Metric {
     fn headers() -> Vec<&'static str> {
         vec!["ID", "NAME", "TYPE", "CREATED"]

src/commands/metrics.rs

@@ -1,7 +1,9 @@
 use anyhow::Result;
 use clap::{Args, Subcommand};
 
-use crate::client::models::{CreateMetricRequest, ListParams, MetricType, UpdateMetricRequest};
+use crate::client::models::{
+    CreateMetricRequest, ListParams, MetricType, TestMetricRequest, UpdateMetricRequest,
+};
 use crate::client::CovalClient;
 use crate::input_json::{self, InputJsonArg};
 use crate::next_actions;
@@ -17,6 +19,7 @@ pub enum MetricCommands {
     Create(Box<CreateArgs>),
     Update(Box<UpdateArgs>),
     Delete(DeleteArgs),
+    Test(TestArgs),
 }
 
 impl MetricCommands {
@@ -28,6 +31,7 @@ impl MetricCommands {
             Self::Create(_) => "create",
             Self::Update(_) => "update",
             Self::Delete(_) => "delete",
+            Self::Test(_) => "test",
         }
     }
 }
@@ -202,6 +206,18 @@ pub struct DeleteArgs {
     metric_id: String,
 }
 
+#[derive(Args)]
+pub struct TestArgs {
+    /// The metric ID to test
+    metric_id: String,
+    /// The simulation output ID to run the metric against
+    #[arg(long)]
+    simulation_output_id: String,
+    /// Optional developer identifier for debugging
+    #[arg(long)]
+    dev_id: Option<String>,
+}
+
 pub async fn execute(cmd: MetricCommands, client: &CovalClient, ctx: &OutputContext) -> Result<()> {
     let operation = cmd.operation();
     match cmd {
@@ -344,6 +360,23 @@ pub async fn execute(cmd: MetricCommands, client: &CovalClient, ctx: &OutputCont
                 next_actions::delete_result("metrics"),
             );
         }
+        MetricCommands::Test(args) => {
+            let req = TestMetricRequest {
+                simulation_output_id: args.simulation_output_id,
+                dev_id: args.dev_id,
+            };
+            let response = client.metrics().test(&args.metric_id, req).await?;
+            emit_one_with_actions(
+                ctx,
+                "metrics",
+                operation,
+                &response,
+                vec![
+                    next_actions::get("metrics", &args.metric_id).primary(),
+                    next_actions::context("metrics"),
+                ],
+            );
+        }
     }
     Ok(())
 }

tests/cli_tests.rs

@@ -3045,6 +3045,71 @@ fn test_metrics_create_composite_metadata_requires_criteria() {
         .stderr(predicate::str::contains("--criteria is required"));
 }
 
+#[tokio::test]
+async fn test_metrics_test_subcommand() {
+    let mock_server = MockServer::start().await;
+
+    Mock::given(method("POST"))
+        .and(path("/v1/metrics/met_abc/test"))
+        .and(header("X-API-Key", "test_key"))
+        .and(body_partial_json(json!({
+            "simulation_output_id": "simout_def456"
+        })))
+        .respond_with(ResponseTemplate::new(202).set_body_json(json!({
+            "metric_output_ulid": "01HXKZ4M5N6P7Q8R9STVWXYZAB"
+        })))
+        .mount(&mock_server)
+        .await;
+
+    coval()
+        .arg("--api-key")
+        .arg("test_key")
+        .arg("--api-url")
+        .arg(mock_server.uri())
+        .arg("metrics")
+        .arg("test")
+        .arg("met_abc")
+        .arg("--simulation-output-id")
+        .arg("simout_def456")
+        .assert()
+        .success()
+        .stdout(predicate::str::contains("01HXKZ4M5N6P7Q8R9STVWXYZAB"));
+}
+
+#[tokio::test]
+async fn test_metrics_test_subcommand_with_dev_id() {
+    let mock_server = MockServer::start().await;
+
+    Mock::given(method("POST"))
+        .and(path("/v1/metrics/met_abc/test"))
+        .and(header("X-API-Key", "test_key"))
+        .and(body_partial_json(json!({
+            "simulation_output_id": "simout_def456",
+            "dev_id": "debug-trace-001"
+        })))
+        .respond_with(ResponseTemplate::new(202).set_body_json(json!({
+            "metric_output_ulid": "01HXKZ4M5N6P7Q8R9STVWXYZAB"
+        })))
+        .mount(&mock_server)
+        .await;
+
+    coval()
+        .arg("--api-key")
+        .arg("test_key")
+        .arg("--api-url")
+        .arg(mock_server.uri())
+        .arg("metrics")
+        .arg("test")
+        .arg("met_abc")
+        .arg("--simulation-output-id")
+        .arg("simout_def456")
+        .arg("--dev-id")
+        .arg("debug-trace-001")
+        .assert()
+        .success()
+        .stdout(predicate::str::contains("01HXKZ4M5N6P7Q8R9STVWXYZAB"));
+}
+
 #[tokio::test]
 async fn test_test_cases_create_with_expected_behaviors() {
     let mock_server = MockServer::start().await;