Unverified50% confidenceOpinionExact time
Forever Security said it reproduced the attack itself, claiming that against Codex the attack file succeeded in planting malware in more than 90% of attempts, and that against Claude Code another planting file had a success rate of about half, but these figures come from the company's own tests.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/31/2026
First Seen
Valid until: 11/14/2026
Sources
AI Weekly: GPT-6.1 Terra Enters at Low Price, AGENTS.md Becomes a New Attack Surface
youtubeAI Without the Noise10/29/2026
Related Entities
Related Claims
UnverifiedOf the three intrusions, one used password brute-force guessing, while the other two involved discovering credentials in public code repositories and using them to access protected systems.64% similarUnverifiedA replay attack is a cybersecurity threat in which an attacker intercepts a legitimate authorization message and retransmits it unchanged at a later time to trick the system into executing the same operation again.61% similarUnverifiedManifold Security disclosed 8 vulnerabilities across 7 AI coding agents including Claude Code, Codex, and Cursor, which attackers can exploit through malicious configurations to induce the agents to execute programs.61% similarUnverifiedThe vulnerability disclosure mechanism in traditional cybersecurity typically gives vendors a 90-day remediation window, after which disclosure is made public regardless of whether the fix has been applied.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1119331API
curl https://kongchang.com/api/v1/knowledge/claims/1119331MCP
get_claim(id=1119331)