diff --git a/CHANGELOG.md b/CHANGELOG.md
index 4b5edce18..42ac39738 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -74,16 +74,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   contains the panic and treats the fragment as plain text, so the request is
   compressed and forwarded normally instead of returning HTTP 500
   ([#1547](https://github.com/headroomlabs-ai/headroom/issues/1547)).
-- `headroom wrap opencode` (and any other `@ai-sdk/anthropic` client) can now
-  use subagents. Mid-turn message coalescing keyed concurrent streaming
-  requests by `md5(model:system[:500])`, so a subagent sharing the main
-  agent's model and system prefix collided with the active stream: its request
-  was answered with a bare `202` and never forwarded, and the stream ended by
-  emitting a non-standard `headroom_pending_messages` SSE event the client
-  couldn't parse (`invalid_union / No matching discriminator`). Coalescing is a
-  Claude Code-only protocol and is now gated to Claude Code clients; other
-  harnesses stream normally
-  ([#1608](https://github.com/headroomlabs-ai/headroom/issues/1608)).
 - Proactive expansion blocks injected into user turns are now wrapped in
   `<headroom_proactive_expansion>` XML tags, giving downstream consumers
   (LLMs, loggers, attribution parsers) a machine-readable provenance