feat: add configurable row limit for AI query tools #8741
Conversation
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
Add AIQueryRowLimit config option (default 500 rows) to prevent AI agents from pulling too much data into context, which can cause context window overflow. Changes: - Add AIQueryRowLimit to InstanceConfig (configurable via rill.ai.query_row_limit) - Update query_metrics_view to cap results at AIQueryRowLimit - Update query_sql to use configurable limit instead of hardcoded 1000 - Update tool specs to document the 500 row limit - Add test for row limit enforcement https://claude.ai/code/session_01NtYNgGmYmZoy9tYya9q5ws
begelundmuller
commented
Feb 2, 2026
runtime/ai/metrics_view_query.go
Outdated
| • 'time_range' is inclusive of start time, exclusive of end time | ||
| • 'time_range.time_dimension' (optional) specifies which time column to filter; defaults to the metrics view's default time column | ||
| • Include 'sort' and 'limit' parameters to optimize query performance and avoid unbounded result sets | ||
| • Results are capped at 500 rows maximum to prevent context overflow; use filters and aggregations to reduce result size |
Contributor
Author
There was a problem hiding this comment.
Load number from instance config instead of hard coding
runtime/ai/metrics_view_query.go
Outdated
|
|
||
| // Cap the limit to prevent AI queries from pulling too many rows | ||
| if cfg.AIQueryRowLimit > 0 { | ||
| args = capLimit(args, cfg.AIQueryRowLimit) |
Contributor
Author
There was a problem hiding this comment.
Pass the limit in the resolver args (args, not props), and consume it in runtime/resolvers/metrics.go resolver implementation instead of having a hacky util function in this file.
runtime/ai/query_sql.go
Outdated
| Name: QuerySQLName, | ||
| Title: "Query SQL", | ||
| Description: "Execute a raw SQL query against an OLAP connector to introspect data.", | ||
| Description: "Execute a raw SQL query against an OLAP connector to introspect data. Results are capped at 500 rows maximum to prevent context overflow; use LIMIT, WHERE, or aggregations to reduce result size.", |
Contributor
Author
There was a problem hiding this comment.
Same comment as for the metrics query tool
runtime/ai/query_sql.go
Outdated
Comment on lines
73
to
77
| // Apply the AI query row limit (default 500) | ||
| limit := cfg.AIQueryRowLimit | ||
| if limit <= 0 { | ||
| limit = 500 // Fallback default | ||
| } |
Contributor
Author
There was a problem hiding this comment.
Don't repeat the default cap here, it's already in the config defaults. Just pass the value through.
- Move limit cap logic from metrics_view_query.go to resolvers/metrics.go - Pass limit_cap via resolver args instead of using hacky util function - Remove hardcoded "500" from tool spec descriptions - Remove fallback default in query_sql.go, just pass config value through https://claude.ai/code/session_01NtYNgGmYmZoy9tYya9q5ws
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.
Adds an AI query row cap config option to prevent AI agents
from pulling too much data into context.
The row limit is configurable with rill.ai.query_row_limit (default 500).