Skip to content

Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors

Rust

test(hook-spool): deflake dead-endpoint drain test under load

AkitaOnRailsd03efc8f7b5e
diff --git a/crates/ai-memory-cli/src/commands/hook_spool.rs b/crates/ai-memory-cli/src/commands/hook_spool.rs
index cd74e2d5..0763f08a 100644
--- a/crates/ai-memory-cli/src/commands/hook_spool.rs
+++ b/crates/ai-memory-cli/src/commands/hook_spool.rs
@@ -1846,11 +1846,17 @@ mod tests {
.unwrap();
}
+ // Generous budgets: the assertion is about ORDERING (a dead head
+ // must not block the deliverable tail), not latency. Tight timeouts
+ // (500ms per event) flaked under `--all-targets` CPU contention when
+ // a live localhost POST briefly exceeded them and a deliverable
+ // entry timed out spuriously. The dead head still connection-refuses
+ // immediately regardless of the ceiling.
let r = drain(
&spool,
tmp.path(),
+ Duration::from_secs(30),
Duration::from_secs(5),
- Duration::from_millis(500),
)
.await;