Warp, the agentic development environment (ADE) built on a modern terminal interface, has disclosed a security incident involving a third-party vendor that exposed agent decision logs and user telemetry data. Founder and CEO Zach Lloyd said the company estimates roughly 1% of Warp’s user base may have been affected, out of a platform reported to have at least 800,000 users.
Warp provides developers with a terminal, code editor, task orchestration, and agent review tools in a single workspace, allowing them to direct AI agents to write, test, and deploy code. According to Lloyd, the company became aware on a Friday of what it believes was an active, narrow-banded exploitation of a vendor in Warp’s infrastructure. The team said it detected the activity within hours and was able to remediate and mitigate the impact.
Warp has not clarified whether the exposed data included source code, prompts, credentials, or other sensitive material, though agent traces and telemetry were identified as potentially affected. Analysis of the incident reportedly points to an attack powered by an agent swarm using what the company described as a very sophisticated penetration and escalation technique.
Response and investigation
Lloyd said Warp has contacted the third-party vendor regarding the vulnerability, though no further details on the vendor or the flaw have been shared publicly. The company has engaged CrowdStrike to assist with the ongoing investigation and has notified law enforcement authorities. Users identified as impacted have been contacted directly.
Warp is advising all users to remain vigilant, monitor their account activity, and treat any unsolicited communications referencing the incident with increased caution.
Part of a broader pattern
The disclosure follows a similar incident at AI coding platform OpenCode, which recently notified users that a breach exposed stored API keys used with external providers. That incident affected about 1.4% of OpenCode users and led to unauthorized charges of up to $1,700 for some customers. Together, the two incidents highlight growing risk around third-party vendor dependencies and supply chain exposure in AI-assisted development tooling, as agentic platforms increasingly handle sensitive telemetry, credentials, and code on behalf of developers.
