Claude Code (stream-json stdin) and Codex (turn/steer pinned to the active
turn) receive a prompt sent while a turn runs immediately. Other harnesses
(OpenCode) get a durable chat_prompt_queue row owned by the session: it
survives the last client detaching and a server restart, is delivered in
order when the turn ends with or without a client, and is deduplicated by
operation id. chat_subscribe carries the queue, chat_queue frames report
delivered/removed/failed changes, and chat_unqueue removes a waiting prompt.
Verify typed ThreadTokenUsage/TokenUsageBreakdown/Thread/Turn against openai/codex@6fb1c3903d. Use native totalTokens and latest-call context tokens without re-adding cached/reasoning subsets. Replay bounded per-connection snapshots and clear on room/source release.
Cost stays unavailable: no subscription USD usage is reported by this notification. Throughput stays unavailable even with native Turn.durationMs because neither last-call nor cumulative-session output is a compatible turn-specific numerator. Thread readback exposes no token usage; no extra reads, resume, or live provider calls added.
Checks: go test -race ./internal/codex; go vet ./internal/codex; go test ./internal/server ./internal/protocol; git diff --check.
Store a chat's harness with its operator session, expose chat_backends and a
backend choice in New chat, and route each session through a manager. The Codex
adapter uses the App Server's Unix WebSocket for threads, turns, history,
questions and interrupt; OpenCode remains the default and Claude is not
advertised until its sidecar and login are wired.