Repository navigation
fix(improve): reserve gateway input within the weekly budget - #332
Merged
Merged
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The weekly KB evolution run stopped after MiMo returned more billable input tokens than the visible request byte estimate: its $0.02645874 actual charge exceeded the $0.0249685625 reservation, although the existing $20 weekly envelope still had $19.52 available. Compatible gateways can add input outside the submitted payload, so visible bytes alone do not bound their input cost.
Add
IMPROVE_INPUT_CONTEXT_LIMITS, an operator-owned map of exact request model names to provider context ceilings. Reservations use the larger of visible bytes and the configured ceiling, priced at the highest configured input/cache rate, plus maximum output cost. Shadow experiments inherit the governor’s exact map and reject overrides. Settlement, the weekly dollar/judge caps, and the existing breach stop remain intact. The operator guide documents MiMo’s 1M context ceiling, audited breach reconciliation, and scheduling after the Monday UTC week reset (09:00 Shanghai time).Validation: 116 passed, 1 skipped across budget, shadow experiments, P2/P3 improvement, evolution, and autonomous evolution suites; 13 targeted regressions passed after adding the parent-versus-child configuration check. The recorded MiMo usage settles under a $0.481152 conservative reservation; a smaller existing cap refuses before dispatch. No paid model calls were used.