OpenCode Console breach: replace provider keys you saved there

If you received an email from OpenCode about a “potential provider key breach”, replace every AI provider key you saved in an OpenCode Console workspace. Do it at each provider’s own dashboard, not through a link in an email. OpenCode says people who got no email were not affected.

What OpenCode has said

On 17 October 2026 at 17:56 UTC, OpenCode posted on X that it “had a security breach last night related to a large scale distillation attack that impacted 1.4% of OpenCode Console users.” It said it had emailed all of them that morning. In an edited version twelve minutes later it added: “If you did not get an email you were not impacted, it was a subset of users who had external providers linked.” The post ends: “The incident is ongoing so we cannot share more details at this point but we are working with Anthropic, OpenAI, and Google to remediate.”

OpenCode Console is the company’s hosted service. Its status page describes the Console API as “Dashboard, workspaces, and keys”. Dax Raad, who builds OpenCode, replied to an Anthropic engineer the same evening: “note this is the OpenCode Console, not OpenCode itself, those keys never come to us.” OpenCode’s statements cover keys saved in hosted Console workspaces. They say nothing about keys kept only on your own machine.

What the email says

The email itself is known from a screenshot a user shared on Reddit, reposted on X and quoted by On The Wire. We have not seen the email directly. As shown in that screenshot, it says OpenCode “identified unauthorized access to a legacy system that we were in the process of decommissioning”, and that third-party provider keys “saved in your workspace may have been exposed.” It asks recipients to “revoke and replace those keys through the relevant provider’s dashboard, update any applications using them, and review recent usage for unexpected activity.”

What Anthropic did

Lydia Hallie, who works on Claude Code at Anthropic, wrote on X on 17 October: “We’ve been working with OpenCode on yesterday’s breach and have revoked Claude API keys we believe were affected.” She added: “If you saved a key in OpenCode and want to be extra careful, please create a new one!”

What to do

1. Revoke at the source. Open each provider’s console yourself (Anthropic, OpenAI, Google and any other provider you linked) and revoke the keys you saved in Console. Deleting a key inside OpenCode does not stop a copy someone already has. Cyera also warns against breach-email links, “since notices like these are common phishing bait.”
2. Issue new keys and update callers. Replace the key in every app, CI secret and script that used it, then confirm the old key fails.
3. Check usage and billing from 16 October onward for models, volumes or hours you do not recognize. Some users on X described unexpected charges. Those are their own reports, not verified, and Dax Raad told two of them to email him with details “and we’ll get you refunded.”
4. Set a hard cap. Anthropic’s API docs say you can set your own monthly spend limit on the Billing page in the Claude Console. Google Cloud supports budgets and budget alerts, and spend-cap budgets for supported services. A cap limits the bill while you rotate keys. It cannot undo a theft.

What is not known

OpenCode has not said who was behind the access, how the legacy system was reached, how the keys were stored, how many accounts 1.4% represents, or what “distillation attack” means here. On 19 October the status page listed no incident for the breach, only an unrelated free-inference slowdown, and the project’s GitHub advisories had nothing newer than 24 September. We found no newer official update on OpenCode’s X account, status page or changelog as of 19 October.

One side effect: on 18 October a Codewhale commit removed README text that promoted storing provider keys in a Codewhale account, citing “OpenCode users’ stored OpenAI and Claude keys” being exposed. We covered Codewhale’s sandbox fixes here, and OpenCode’s local agent here.