Hitonabi/mission-control-v2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(2.0): W1 — fast-Spur ohne Thinking (flotte Antworten) + tool-call-Cap

Browse Source 
Gateway injiziert chat_template_kwargs.enable_thinking=false fuer die fast-
Spur (Qwen3.6 ist Reasoning-Modell → sonst lahm/leer). heavy behaelt
Thinking. Env MC_FAST_NO_THINK. Hermes-Thrash war poisoned Persistent-
Session (fresh=clean, 34k statt 249k verifiziert); code_execution.max_tool_
calls 50->20 auf der Box (Historie unangetastet).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Hitonabi
2026-06-25 14:24:57 +02:00
parent 77b6dee02f
commit 501ba36b89
3 changed files with 9 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/services/router_logic.py
+3
View File
@@ -9,6 +9,9 @@ import re
FAST = os.environ.get("MC_ROUTE_FAST", "fast")
HEAVY = os.environ.get("MC_ROUTE_HEAVY", "heavy")
HEAVY_CHARS = int(os.environ.get("MC_GATEWAY_HEAVY_CHARS", "8000"))
# Thinking auf der fast-Spur ausschalten → flotte Alltags-Antworten (Qwen3.6 ist ein
# Reasoning-Modell). heavy behält Thinking für harte Aufgaben. Abschaltbar via Env.
FAST_NO_THINK = os.environ.get("MC_FAST_NO_THINK", "1") not in ("0", "false", "")
_HEAVY_KW = re.compile(
r"\b(beweis|prove|theorem|refactor|architect|komplex|complex|schwierig|"