[kzahel]

Can't Opencode just modify their implementation to use the anthropic claude code SDK directly? The issue is they were spoofing oauth. I tried OpenCode before this whole drama and immediately noticed the oauth spoofing and never authorized it. Doesn't opencode speak ACP? https://agentclientprotocol.com/overview/agents

[dd8601fn]

It already does.

You can use the Anthropic API in any tool, but these users wanted to use the claude code subscription.

[kzahel]

OpenCode wasn't using claude CLI at all (or claude SDK). They were using their own agent loop and bypassing claude cli entirely (except for spoofing auth).

[macinjosh]

The SDK bundles Claude code and uses it for its agentic work. The SDK really only lets you control the UI layer. It als doesn’t yet fully support plan mode.

[kzahel]

I use the SDK in my app and it works fine with plan mode. I don't deal with auth at all. I detect if the CLI is installed and it just reuses whatever auth the user has already setup. Works fine.

[nmfisher]

> I detect if the CLI is installed and it just reuses whatever auth the user has already setup.

Isn't this what they just explicitly banned?

[kzahel]

no, they banned use of the model without the CLI harness/SDK when using the subscription plans. Opencode was spoofing requests as if they were coming from claude code CLI, and controlling the agent loop / tool call totally internally. Anthropic wants subscription plans to use the CLI/SDK.