Repository navigation
fix(gov): enforce daily index call limits - #5144
Merged
Merged
Conversation
The gov index is free and reachable as an Alexandria provider that the Exchange calls in-cluster, past the per-minute limiters, with up to 10 calls per request. One team ran about 223k gov searches in a day and saturated the search-index reranker. Every path into a free index now spends from one per-team budget per UTC day, one unit per call, so a 10-call batch spends 10: the Alexandria provider, /v2/search/gov and the gov category on /v2/search. A request that does not fit is refused whole with 429 and Retry-After, before any quote or execution, and spends nothing. Redis failures fail open. Gov is limited at GOV_INDEX_CALLS_PER_DAY (default 500). The research index is wired the same way but stays unlimited until RESEARCH_INDEX_CALLS_PER_DAY is set. Also drops the unused isPale export flagged by knip.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
All reported issues were addressed across 7 files
This PR changes payment, billing, or quota logic. Ultrareviews find 2.4x more serious bugs than standard reviews. Comment @cubic-dev-ai ultrareview to run one.
View guided diff | Turn on auto-fix | Re-trigger cubic
…alues - Alexandria: index calls reserved before the quote and credit hold are given back on every refusal before execution (and on an exception), so transient quote or billing failures no longer burn a team's daily budget. Refusals after the Exchange executed still count. - Alexandria: send the Retry-After header when a refusal carries retryAfterSeconds, as the other gov paths do. - /v2/search: reserve the gov call immediately before executeSearch, after every rejecting validation, so a 403 spends nothing. - Config: blank GOV_INDEX_CALLS_PER_DAY falls back to 500 and a blank RESEARCH_INDEX_CALLS_PER_DAY stays unset, instead of coercing to 0. - Bound the Redis wait at 500 ms: an unreachable Redis queues commands rather than failing, which broke the fail-open promise. - Tests: release and rollback, Retry-After on the Alexandria route, the /v2/search/gov 429, gov-only counting in retrieve, and gov-search tests no longer depend on a live Redis.
- /v2/search/gov reserves after the request is logged, like /v2/search, so a failed log spends nothing, and gives the call back when the index is not configured. - The gov daily-limit 429 on /v2/search and /v2/search/gov carries code INDEX_DAILY_LIMIT, so clients can tell it from the per-minute limiter and stop retrying until the reset (Alexandria already returns index_daily_limit). - Correct the config comment: gov defaults to 500; only research is unlimited while unset.
Contributor
There was a problem hiding this comment.
All reported issues were addressed across 9 files (changes from recent commits).
Tip: Get a Slack DM when cubic finishes reviewing your PRs. Connect Slack
View guided diff | Turn on auto-fix | Re-trigger cubic
- Alexandria: set Retry-After only for the index_daily_limit refusal. The Exchange already returns retryAfterSeconds for other providers' refusals, which keep their existing responses. - After a Redis timeout the reservation command can still land once Redis reconnects. Its release now waits for that command in the background and gives back what it took, so a request refused before it ran spends nothing. A synchronous eval throw is also caught.
Contributor
There was a problem hiding this comment.
All reported issues were addressed across 11 files (changes from recent commits).
Requires human review: Auto-approval blocked because this review re-detected 1 unresolved P0–P2 issue already reported by Cubic.
Tip: Review your code locally with the cubic CLI to iterate faster.
View guided diff | Turn on auto-fix | Re-trigger cubic
Contributor
There was a problem hiding this comment.
0 issues found across 9 files (changes from recent commits).
Confidence score: 5/5
- Automated review surfaced no issues in the provided summaries.
- No files require special attention.
Requires human review: Auto-approval blocked by 1 unresolved P0–P2 issue from previous reviews.
View guided diff | Turn on auto-fix | Re-trigger cubic
Restrict daily quotas to government index requests and leave other indexes unchanged. Make reservation refunds idempotent and register the API-level quota error.
capy-ai
Bot
changed the title
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.
Limit government-index calls per team per UTC day across
/v2/search/gov, gov-category search, and the government Alexandria provider. Non-gov requests retain their existing behavior.GOV_INDEX_CALLS_PER_DAY(default: 500).Retry-After.Validation: API build, formatting, and targeted local unit/regression checks. Additional validation tests remain local; the PR contains production changes only.