fix: add import alias tracking and getattr detection to anti-hack#37
Merged
Conversation
…er 1 Two new defenses: 1. Import alias resolution: 'import torch as tr' followed by 'tr.sum()' is now correctly identified as torch.sum() 2. getattr detection: 'getattr(torch, "sum")' is caught as dynamic torch API access Also detects combinations of both (alias + getattr). All three detection vectors correctly handle the torch API whitelist: - Direct call: torch.sum() -> check whitelist - Alias call: tr.sum() -> resolve alias -> check whitelist - Dynamic call: getattr(torch, "sum") -> check whitelist Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Close two anti-hack bypass vectors:
import torch as tr; tr.sum()now correctly detected as forbidden torch APIgetattr(torch, "sum")now caughtAlias resolution + getattr detection work together for combined attacks.
Both respect the same torch API whitelist.