Add Docs for ranking_instructions in Query Agent Search Mode

docs/query-agent/_includes/code/search_mode.mts

@@ -102,4 +102,15 @@ for (const obj of filteringResponse.searchResults.objects) {
 }
 // END FilteringExample
 
+// START RankingInstructions
+const rankingResponse = await qa.search("wireless headphones", {
+    rankingInstructions: "Prioritize products with recent reviews and in-stock availability over price.",
+    limit: 10,
+});
+
+for (const obj of rankingResponse.searchResults.objects) {
+    console.log(`Product: ${obj.properties['name']} - $${obj.properties['price']}`);
+}
+// END RankingInstructions
+
 await client.close();

docs/query-agent/_includes/code/search_mode.py

@@ -107,6 +107,17 @@
     print(f"Product: {obj.properties['name']} - ${obj.properties['price']}")
 # END FilteringExample
 
+# START RankingInstructions
+search_response = qa.search(
+    "wireless headphones",
+    ranking_instructions="Prioritize products with recent reviews and in-stock availability over price.",
+    limit=10,
+)
+
+for obj in search_response.search_results.objects:
+    print(f"Product: {obj.properties['name']} - ${obj.properties['price']}")
+# END RankingInstructions
+
 # --- Async code examples in string as top-level await doesn't work, full code will be executed in 
 # asyncio.run below
 

docs/query-agent/guides/search_mode.md

@@ -69,6 +69,7 @@ The `.search()` method accepts several arguments:
 | `limit` | `int` | The maximum number of results returned in this page of results. Defaults to `20`. Use [`.next()`](#pagination) to fetch additional pages. |
 | `filtering` | `Literal["recall", "precision"]` | Either `"recall"` or `"precision"` to control filter generation. `"recall"` favors more results across filter interpretations; `"precision"` favors strict intent match. See [Customized filtering](#customized-filtering) below. |
 | `diversity_weight` | `float \| None` | A value between `0.0` and `1.0` that biases the result ranking towards diversity using Maximal Marginal Relevance (MMR). See [Diversity ranking](#diversity-ranking) below. |
+| `ranking_instructions` | `str \| None` | Natural-language instructions that guide the reranker on which aspects of relevance to prioritize (e.g., `"Prioritize products with recent reviews and in-stock availability over price."`). Only affects the ordering of the retrieved results, not which results are retrieved. Limited to 500 tokens. See [Ranking instructions](#ranking-instructions) below. |
 
 </TabItem>
 <TabItem value="ts_agents" label="JavaScript/TypeScript">
@@ -79,6 +80,7 @@ The `.search()` method accepts several arguments:
 | `limit` | `number` | The maximum number of results returned in this page of results. Defaults to `20`. Use [`.next()`](#pagination) to fetch additional pages. |
 | `filtering` | `"recall" \| "precision"` | Either `"recall"` or `"precision"` to control filter generation. `"recall"` favors more results across filter interpretations; `"precision"` favors strict intent match. See [Customized filtering](#customized-filtering) below. |
 | `diversityWeight` | `number` | A value between `0.0` and `1.0` that biases the result ranking towards diversity using Maximal Marginal Relevance (MMR). See [Diversity ranking](#diversity-ranking) below. |
+| `rankingInstructions` | `string` | Natural-language instructions that guide the reranker on which aspects of relevance to prioritize (e.g., `"Prioritize products with recent reviews and in-stock availability over price."`). Only affects the ordering of the retrieved results, not which results are retrieved. Limited to 500 tokens. See [Ranking instructions](#ranking-instructions) below. |
 
 </TabItem>
 </Tabs>
@@ -138,6 +140,35 @@ To use diversity ranking with target vectors, set the single target vector you w
     </TabItem>
 </Tabs>
 
+### Ranking instructions
+
+Search Mode uses reranking to improve the relevancy of results. You can pass natural-language ranking instructions to guide the instruction-following reranker on which aspects of relevance to prioritize, for example, recency or availability.
+
+Ranking instructions are reused when [paginating](#pagination) with `.next()`, so the ordering stays consistent across pages.
+
+:::note Length limit
+Ranking instructions are limited to 500 tokens. Requests with longer instructions fail validation.
+:::
+
+<Tabs className="code" groupId="languages">
+    <TabItem value="py_agents" label="Python">
+        <FilteredTextBlock
+            text={PyCode}
+            startMarker="# START RankingInstructions"
+            endMarker="# END RankingInstructions"
+            language="py"
+        />
+    </TabItem>
+    <TabItem value="ts_agents" label="JavaScript/TypeScript">
+        <FilteredTextBlock
+            text={TSCode}
+            startMarker="// START RankingInstructions"
+            endMarker="// END RankingInstructions"
+            language="ts"
+        />
+    </TabItem>
+</Tabs>
+
 ## Response
 The Search Mode response has the following properties: