@@ -28,7 +28,7 @@ The agent for asking questions about your data. Truly read-only mode for safe da
 - **Bash**: Safe commands auto-allowed (`ls`, `grep`, `cat`, `head`, `tail`, `find`, `wc`). dbt read commands allowed (`dbt list`, `dbt ls`, `dbt debug`). Everything else denied.
 - **Web**: Fetch and search allowed without prompts
 - **Schema/warehouse/finops**: All inspection tools available
-- **Workspace knowledge**: Loads workspace and project skills; reads and reloads workspace memory. Saving memory is done with `builder`.
+- **Workspace knowledge**: Loads workspace and project skills; reads workspace memory and views this checkout's memory audit log. Reloading memory syncs the local copy with the workspace, which can remove a local copy of a block the workspace archived. Saving memory is done with `builder`.
 
 !!! tip
     Use `analyst` when exploring data to ensure no accidental writes. Switch to `builder` when you're ready to create or modify models.
 
@@ -134,7 +134,7 @@ Estimated savings: ~60 credits/week (71% reduction on this query alone)
 - Search warehouse metadata
 - Detect PII columns
 - Save training entries
-- Read and reload workspace memory, and answer questions about team conventions from workspace skills (saving memory is done with builder)
+- Read and reload workspace memory, view this checkout's memory audit log, and answer questions about team conventions from workspace skills (saving memory is done with builder)
 
 ---
 
 
@@ -630,9 +630,10 @@ for (const turn of [
   { agent: "analyst", tools: undefined, expected: { read: true, write: false } },
   { agent: "builder", tools: undefined, expected: { read: true, write: true } },
   { agent: "builder", tools: { altimate_memory_read: false }, expected: { read: false, write: true } },
+  { agent: "analyst", tools: { altimate_memory_read: false }, expected: { read: false, write: false } },
 ] as const) {
   noLLMServer.instance(
-    `prompt loop describes only the memory tools ${turn.agent} keeps${turn.tools ? " with read switched off on the request" : ""}`,
+    `prompt loop passes the identity section only the memory tools ${turn.agent} keeps${turn.tools ? " with read switched off on the request" : ""}`,
     () => Effect.gen(function* () {
       const section = spyOn(WorkspaceIdentity, "systemSection").mockResolvedValue("")
       const sent: Pick<LLM.StreamInput, "tools" | "agent" | "user">[] = []