Skip to main content
Memory Router returns diagnostic headers so you can see exactly how each request was handled. Every response tells you what the Router did — what context it changed, how many tokens it saved, and which memories it touched.

​
Diagnostic Response Headers

Each response carries the following diagnostics:
SignalWhat it tells you
Conversation IDThe thread the request was attributed to, so you can group and inspect turns.
Context modifiedWhether memory was injected or history was trimmed for this call.
Token countsHow many tokens were sent after optimization versus a raw replay.
Memories touchedHow many memory chunks were retrieved and how many were created.
Use Token counts to verify savings as conversations grow, and Context modified to confirm that memory injection is happening on the turns you expect.

​
Graceful Fallback

In BYOK mode, Memory Router fails open: if MemoryLake is ever unavailable, the request passes straight through to your provider so your application keeps working with zero downtime.
Because the Router speaks your provider’s protocol, a passthrough request returns the same response shape as a direct call — your application keeps working without code changes.

​
Next Steps

Deployment Modes

BYOK vs. hosted, base URLs, and supported providers.

FAQ

Common questions about code, providers, security, and pricing.