From fc4080c58a4deef04e49d5ead4ccffc5d7170a68 Mon Sep 17 00:00:00 2001
From: teknium1 <teknium1@gmail.com>
Date: Mon, 16 Mar 2026 07:34:29 -0700
Subject: [PATCH] fix(cli): add <THINKING> to streaming tag suppression list

Anthropic native models emit <THINKING> tags in text content (separate
from the SDK's thinking_delta events). Without suppression, these tags
leak into the streamed CLI output. Found during live provider testing.
---
 cli.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/cli.py b/cli.py
index b7203e9063..2aa181797a 100755
--- a/cli.py
+++ b/cli.py
@@ -1431,8 +1431,8 @@ class HermesCLI:
         # These tags are model-generated (system prompt tells the model
         # to use them) and get stripped from final_response. We must
         # suppress them during streaming too.
-        _OPEN_TAGS = ("<REASONING_SCRATCHPAD>", "<think>", "<reasoning>")
-        _CLOSE_TAGS = ("</REASONING_SCRATCHPAD>", "</think>", "</reasoning>")
+        _OPEN_TAGS = ("<REASONING_SCRATCHPAD>", "<think>", "<reasoning>", "<THINKING>")
+        _CLOSE_TAGS = ("</REASONING_SCRATCHPAD>", "</think>", "</reasoning>", "</THINKING>")
 
         # Append to a pre-filter buffer first
         self._stream_prefilt = getattr(self, "_stream_prefilt", "") + text