(skywalking-query-protocol) 01/01: ai-agent: filter the conversation list by conversation id and by title

[wu...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

wu-sheng pushed a commit to branch ai-agent-list-filters
in repository https://gitbox.apache.org/repos/asf/skywalking-query-protocol.git

commit 66edb7a5405b92a623b547e3b253af4324342a02
Author: Wu Sheng <wu...@foxmail.com>
AuthorDate: Sun Sep 6 10:53:31 2026 +0800

    ai-agent: filter the conversation list by conversation id and by title
    
    Two optional fields on `ConversationListCondition`:
    
    - `conversation: ID` narrows the read to one conversation. It is a storage
      lookup, so the round budget (`limit`) is spent on that conversation alone.
    - `title: String` keeps only the rows whose title contains the text,
      case-insensitively. The title is the newest round's, so the match runs after
      folding and cannot widen the rounds read.
---
 ai-agent-conversation.graphqls | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/ai-agent-conversation.graphqls b/ai-agent-conversation.graphqls
index fac7d85..475e470 100644
--- a/ai-agent-conversation.graphqls
+++ b/ai-agent-conversation.graphqls
@@ -78,6 +78,11 @@ input ConversationListCondition {
     service: ServiceCondition!
     # Optional. Only the conversations pushed by this sender.
     instance: InstanceCondition
+    # Optional. Only this conversation, by id: a storage lookup, so the round budget is spent on it alone.
+    conversation: ID
+    # Optional. Only the conversations whose title contains this text, case-insensitively. The title is the newest
+    # round's, so the match runs after folding and cannot widen the rounds read.
+    title: String
     # The rounds are read newest first, at most this many, then folded to one row per conversation.
     # Default 1000. A page therefore holds at most this many rows.
     limit: Int