AIKIDO-2026-250781
@milkdown/crepe is vulnerable to Regular Expression Denial of Service (ReDoS)
28
Low Risk
This Affects:
@milkdown/crepeTL;DR
The built-in OpenAI and Anthropic LLM providers in the crepe editor normalize a caller-supplied provider baseURL with a regular expression (replace(/\/+$/, '')) that strips trailing slashes. On inputs containing long runs of / that are not anchored at the very end, the \/+$ sub-expression backtracks quadratically, so an application that forwards untrusted input into the provider baseURL can be driven into O(n^2) CPU time and stalled. The regex was flagged by CodeQL as js/polynomial-redos. The fix replaces it with a linear stripTrailingSlashes helper that scans characters via charCodeAt without backtracking.
Who does this affect?
You are affected if you are using a version that falls within the vulnerable range and you enable the crepe AI feature with an OpenAI or Anthropic provider whose baseURL can be influenced by untrusted input.
Background info
@milkdown/crepe is vulnerable to Regular Expression Denial of Service (ReDoS) in versions 7.21.0 - 7.21.0.
How to fix this
Upgrade the @milkdown/crepe library to the patch version.
Are You Affected?
Connect your repositories to instantly see whether vulnerable or malicious packages exist in your codebase.
Free. No credit card required.
SOC 2Compliant
ISO 27001Compliant