Repository navigation
Conversation
Added detailed README for CVE-2026-108592, including vulnerability summary, proof-of-concept, and remediation steps.
Implement a proof of concept for CVE-2026-108592 using a Bubblewrap environment.
Updated README to reflect correct CVE details and vulnerability class.
Added details about the impact of prompt injection attacks on environment variables and their potential exfiltration over the network.
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
CVE ID
CVE-2026-108592
Checklist
2026/CVE-2026-108592README.mdcompleted using the template (CVSS, affected versions, references)Summary
CVE-2026-108592 describes a potential security issue in which commands running inside a Bubblewrap sandbox may inherit environment variables from the host environment (CWE-526). Environment variables sometimes contain sensitive information, such as API keys, access tokens, and service credentials. If a sandbox exposes these values to commands that should not have access to them, sensitive information could be at risk.
Attackers using prompt injection in processed task content can make the agent read API keys from the environment and exfiltrate them over the shared network.
This PoC uses a harmless dummy environment variable to test whether a value defined in the host environment is visible inside the sandbox.
References