diff --git a/sds/sds_utils.nim b/sds/sds_utils.nim
index 87fb50a..445dd35 100644
--- a/sds/sds_utils.nim
+++ b/sds/sds_utils.nim
@@ -328,15 +328,6 @@ proc getRecentHistoryEntries*(
         if not rm.onRetrievalHint.isNil():
           {.cast(raises: []).}:
             entry.retrievalHint = rm.onRetrievalHint(msgId)
-          if entry.retrievalHint.len > 0:
-            # Phase 2B: best-effort hint persistence via V2. Non-fatal —
-            # hints are an optimisation; a missing hint just means the
-            # peer falls back to slower retrieval.
-            let hintRes =
-              await rm.persistence.setRetrievalHint(msgId, entry.retrievalHint)
-            if hintRes.isErr():
-              warn "retrieval hint save failed; continuing",
-                msgId = msgId, detail = hintRes.error
         entry.senderId = channel.messageHistory[msgId].senderId
         entries.add(entry)
       ok(entries)
diff --git a/sds/types/persistence.nim b/sds/types/persistence.nim
index 6f5e459..b077e3f 100644
--- a/sds/types/persistence.nim
+++ b/sds/types/persistence.nim
@@ -27,7 +27,6 @@ export results, sds_message_id, channel_meta, history_update
 type Persistence* = object
   ## Pluggable durability backend. Supplied at `newReliabilityManager`
   ## construction time; defaults to `noOpPersistence()` when not given.
-
   saveChannelMeta*: proc(
     channelId: SdsChannelID, meta: ChannelMeta
   ): Future[Result[void, string]] {.async: (raises: []), gcsafe.}
@@ -42,30 +41,21 @@ type Persistence* = object
     ## maxMessageHistory cap. Callers SHOULD skip this call entirely when
     ## `update.isEmpty`.
 
-  loadChannel*: proc(
-    channelId: SdsChannelID
-  ): Future[Result[ChannelData, string]] {.async: (raises: []), gcsafe.}
+  loadChannel*: proc(channelId: SdsChannelID): Future[Result[ChannelData, string]] {.
+    async: (raises: []), gcsafe
+  .}
     ## Bootstrap on `getOrCreateChannel`. Returns the full prior state, or
     ## an empty `ChannelData` if the channel is new on disk. Failure
     ## propagates to the caller — bootstrap is a durability-intent op.
 
-  dropChannel*: proc(
-    channelId: SdsChannelID
-  ): Future[Result[void, string]] {.async: (raises: []), gcsafe.}
+  dropChannel*: proc(channelId: SdsChannelID): Future[Result[void, string]] {.
+    async: (raises: []), gcsafe
+  .}
     ## Wipe all persisted state for a channel. Called by `removeChannel` /
     ## `resetReliabilityManager`. Backends SHOULD execute atomically.
     ## Failure propagates to the caller — the caller asked us to confirm a
     ## disk wipe and we cannot silently lie.
 
-  setRetrievalHint*: proc(
-    msgId: SdsMessageID, hint: seq[byte]
-  ): Future[Result[void, string]] {.async: (raises: []), gcsafe.}
-    ## Record a retrieval hint for a message id. Called from
-    ## `getRecentHistoryEntries` when an application-supplied hint
-    ## provider returns a non-empty hint. Out-of-band from the
-    ## snapshot/history write path because hints are populated lazily
-    ## during read. Non-fatal on failure.
-
 proc noOpPersistence*(): Persistence =
   ## Default backend: discards all writes, returns an empty snapshot on
   ## load. Used when no real backend is supplied (existing tests and
@@ -87,8 +77,4 @@ proc noOpPersistence*(): Persistence =
         channelId: SdsChannelID
     ): Future[Result[void, string]] {.async: (raises: []).} =
       ok(),
-    setRetrievalHint: proc(
-        msgId: SdsMessageID, hint: seq[byte]
-    ): Future[Result[void, string]] {.async: (raises: []).} =
-      ok(),
   )
diff --git a/tests/in_memory_persistence.nim b/tests/in_memory_persistence.nim
index 087545e..17a4775 100644
--- a/tests/in_memory_persistence.nim
+++ b/tests/in_memory_persistence.nim
@@ -6,7 +6,7 @@
 ##
 ## `failingOps` injects backend failures. Op names match the `Persistence`
 ## field names: "saveChannelMeta", "updateHistory", "loadChannel",
-## "dropChannel", "setRetrievalHint".
+## "dropChannel".
 
 import std/[tables, sets]
 import chronos
@@ -15,7 +15,6 @@ import sds
 type InMemoryStore* = ref object
   lamports*: Table[SdsChannelID, int64]
   log*: Table[SdsChannelID, OrderedTable[SdsMessageID, SdsMessage]]
-  hints*: Table[SdsMessageID, seq[byte]]
   outgoing*: Table[SdsChannelID, OrderedTable[SdsMessageID, UnacknowledgedMessage]]
   incoming*: Table[SdsChannelID, OrderedTable[SdsMessageID, IncomingMessage]]
   outgoingRepair*: Table[SdsChannelID, OrderedTable[SdsMessageID, OutgoingRepairEntry]]
@@ -23,8 +22,7 @@ type InMemoryStore* = ref object
   dropChannelCalls*: Table[SdsChannelID, int]
     ## Per-channel counter; lets tests assert dropChannel is invoked
     ## exactly once per logical drop.
-  failingOps*: HashSet[string]
-    ## Op names that should return an injected backend error.
+  failingOps*: HashSet[string] ## Op names that should return an injected backend error.
 
 proc newInMemoryStore*(): InMemoryStore =
   InMemoryStore(failingOps: initHashSet[string]())
@@ -48,8 +46,7 @@ proc newInMemoryPersistence*(store: InMemoryStore): Persistence =
           store.outgoing[channelId][u.message.messageId] = u
 
         # Incoming buffer.
-        store.incoming[channelId] =
-          initOrderedTable[SdsMessageID, IncomingMessage]()
+        store.incoming[channelId] = initOrderedTable[SdsMessageID, IncomingMessage]()
         for m in meta.incomingBuffer:
           store.incoming[channelId][m.message.messageId] = m
 
@@ -120,11 +117,4 @@ proc newInMemoryPersistence*(store: InMemoryStore): Persistence =
         store.dropChannelCalls[channelId] =
           store.dropChannelCalls.getOrDefault(channelId) + 1
       ok(),
-    setRetrievalHint: proc(
-        msgId: SdsMessageID, hint: seq[byte]
-    ): Future[Result[void, string]] {.async: (raises: []).} =
-      if "setRetrievalHint" in store.failingOps:
-        return err("injected backend failure: setRetrievalHint")
-      store.hints[msgId] = hint
-      ok(),
   )