Repository navigation
🧮 fix: Keep Run Code Spreadsheets Off the File Search Queue - #102
Merged
Merged
Conversation
Programmatic tool calls embedded a 50,000-row CSV in File Search before the code could run, because a spreadsheet routed to Run Code was also queued for File Search. Under automatic reading, a spreadsheet sent to Run Code is now queued for Run Code alone, and returns to the File Search queue only if Run Code does not load.
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the
 |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Under
llmDeliveryPolicy: automatic(#101), a spreadsheet attached to an agent that runs both Run Code and File Search was queued for File Search as well as Run Code. When the model then ran code through programmatic tool calling (run_tools_with_bash), the provisioning callback embedded every queued File Search file first. On apro-sandbox an 8.7 MB, 50,000-row CSV became 50,000 chunks, and the turn hung for over 11 minutes before any code ran.A spreadsheet that automatic reading sends to Run Code is now Run Code's alone: it is not queued for File Search and not listed in the File Search note. If Run Code then fails to load, it goes back on the File Search queue. Classic endpoints, chosen destinations, the agent's own files, per-type
defaultLLMDeliveryPathoverrides and documents moved for size are unchanged. The resolver and initialization share one predicate,isRunCodeSpreadsheet, so they cannot disagree.Type of change
Testing
Automated tests:
packages/data-providerresolver and file-config specs: 385 passed (newisRunCodeSpreadsheetcases).packages/apiinitialize, attachments, client, delivery, encode, resources, fallback, documents and provision specs: 492 passed. New: spreadsheet off the queue and the note, a PDF still queued, classic unchanged, agent-scoped spreadsheet kept, restore when Run Code does not load.api/BaseClient, agents client and Files process specs: 541 passed.tsc --noEmit(data-provider, packages/api) and eslint: clean.Screenshots / recordings
No user-facing change.
Risk / compatibility
llmDeliveryPolicy: automatic.