T3 Code voice cleanup — real server HTTP recording Commit: 392526768806ab7b36854ad9a401955c4bec6a85 Environment: Linux; fresh isolated database; T3CODE_DEV_SEED=0; loopback HTTP. Project name: Server; personal dictionary: []; project dictionary: []. Client vocabulary helper input: {"personal": [], "project": [], "hints": ["", "Server"]} Actual vocabulary helper output: [{"term": "Server", "aliases": []}] CASE: project hint POST /api/voice/post-process Request: {"transcript": "check which services are running on the server and how server performance looks", "draft": {"text": "", "selection": {"start": 0, "end": 0}}, "options": {"speechCustomWords": [{"term": "Server", "aliases": []}], "speechPostProcessingPrompt": {"mode": "default", "customInstructions": ""}}} Actual cleanup-model reply: {"transcript": "check which services are running on the server and how server performance looks", "transcription": "Check which services are running on the server and how server performance looks?"} HTTP status: 200 Actual T3 server response: {"text": "Check which services are running on the Server and how Server performance looks?"} CASE: no custom words control POST /api/voice/post-process Request: {"transcript": "check which services are running on the server and how server performance looks", "draft": {"text": "", "selection": {"start": 0, "end": 0}}, "options": {"speechCustomWords": [], "speechPostProcessingPrompt": {"mode": "default", "customInstructions": ""}}} Actual cleanup-model reply: {"transcript": "check which services are running on the server and how server performance looks", "transcription": "Check which services are running on the server and how server performance looks?"} HTTP status: 200 Actual T3 server response: {"text": "Check which services are running on the server and how server performance looks?"} CASE: explicit custom word control POST /api/voice/post-process Request: {"transcript": "check which services are running on the server and how server performance looks", "draft": {"text": "", "selection": {"start": 0, "end": 0}}, "options": {"speechCustomWords": [{"term": "Server", "aliases": []}], "speechPostProcessingPrompt": {"mode": "default", "customInstructions": ""}}} Actual cleanup-model reply: {"transcript": "check which services are running on the server and how server performance looks", "transcription": "Check which services are running on the server and how server performance looks?"} HTTP status: 200 Actual T3 server response: {"text": "Check which services are running on the Server and how Server performance looks?"}