Skip to content

Feat/eng 648 enable router model #273

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
ea-rus wants to merge 4 commits into
base: staging
Choose a base branch
from
+3 −3
Draft

Feat/eng 648 enable router model #273

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
13d90b5
test with router model
ea-rus Jul 24, 2026
01e7b24
test fail
ea-rus Jul 24, 2026
6fd85c7
rollback
ea-rus Jul 24, 2026
46bd5ae
rollback
ea-rus Jul 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions anton/core/settings.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,9 +9,9 @@ class CoreSettings(BaseSettings):
# turn first hits the router model, which either answers trivial/from-
# context requests directly or delegates to the planning model (optionally
# preloading skills). The mechanism is the "thalamus" (see
# anton/core/llm/thalamus.py); the user-facing knobs stay "router". Off by
# default until evaluated; flip with ANTON_ROUTER_ENABLED=true.
router_enabled: bool = False
# anton/core/llm/thalamus.py); the user-facing knobs stay "router". On by
# default; disable with ANTON_ROUTER_ENABLED=false.
router_enabled: bool = True
# Output budget for the gating call. Deliberately small: a direct
# answer that doesn't fit here is evidence the turn wasn't trivial,
# and the router treats truncation as "delegate".
Expand Down
Loading