Hmbown/CodeWhale · info

recorded requests

Error message

recorded requests

What it means

This .expect("recorded requests") is in the same test in crates/tui/src/client.rs:7687, unwrapping wiremock's MockServer::received_requests(). It panics if the mock server fails to return its recorded request log. The declaration `let requests = ...` is the variable the expect message is attached to; the app-server references shown are an unrelated same-named fixture elsewhere.

Solutions

1. Ensure the MockServer outlives all awaits that use it (bind it to a local variable in the test, not a dropped scope).
2. Keep the assertion order: query received_requests() before any code that could abort the runtime.
3. Pin/upgrade wiremock to a compatible version and re-run the test.
4. If the server genuinely recorded nothing, note this test expects an EMPTY list — an Err here is still a harness failure, distinct from the is_empty assertion below it.

Example fix

// before
let requests = server.received_requests().await.expect("recorded requests");
// after
let requests = server
    .received_requests()
    .await
    .unwrap_or_default();
assert!(requests.is_empty(), "Concentrate must not treat unauthenticated GET /v1/models as key verification");

// tolerate a missing recording log in tests
let requests = server.received_requests().await.unwrap_or_default();
assert!(requests.is_empty(), "expected no requests, got {}", requests.len());

When it happens

Trigger: server.received_requests().await returning Err — happens when the wiremock server has shut down or its internal recording channel failed by the time the test queries it; typically a test-ordering/lifecycle problem rather than a client bug.

Common situations: Server dropped or tokio runtime torn down before received_requests() is awaited; mock server started but a panic elsewhere in the test left it in a bad state; wiremock version incompatibilities in the recording API.

Related errors

• recorded request
• health check
• Codewhale chat request should succeed
• Codewhale client should resolve its model route
• Concentrate catalog delta

AI-assisted analysis of Hmbown/CodeWhale@73e0f67d83 (2026-09-22). Data as JSON: /api/errors/9cfad1aab19d9bf5. Report an issue: GitHub.

Appendix: source

Thrown at crates/tui/src/client.rs:7687

                expected,
                "HTTP {status}: {message}"
            );
        }
    }

    /// Concentrate's `GET /v1/models` is unauthenticated, so a 2xx must not
    /// count as key verification. Guided setup treats the probe as unobserved;
    /// health_check must not issue the request either.
    #[tokio::test]
    async fn concentrate_health_check_does_not_treat_unauthenticated_models_as_key_proof() {
        let server = MockServer::start().await;
        let client = concentrate_client(&server, DEFAULT_CONCENTRATE_MODEL);

        assert!(client.health_check().await.expect("health check"));
        assert!(!provider_api_key_verification_is_observed(
            ApiProvider::Concentrate
        ));
        let requests = server.received_requests().await.expect("recorded requests");
        assert!(
            requests.is_empty(),
            "Concentrate must not treat unauthenticated GET /v1/models as key verification"
        );
    }

    /// `GET /v1/models` needs no key and answers the OpenAI list shape; rows
    /// are provider-scoped, the default is marked, and unknowns stay unclaimed.
    /// https://concentrate.ai/docs/api-reference/endpoint/list-models
    #[tokio::test]
    async fn concentrate_live_catalog_is_provider_scoped_and_marks_the_default() {
        let server = MockServer::start().await;
        Mock::given(method("GET"))
            .and(path("/v1/models"))
            .respond_with(ResponseTemplate::new(200).set_body_json(json!({
                "object": "list",
                "data": [
                    {"id": "claude-fable-5", "object": "model", "owned_by": "anthropic"},

View on GitHub (pinned to 73e0f67d83)