Skip to content

Releases

v0.2.0

Choose a tag to compare

@Doctorthe113 Doctorthe113 released this 15 Nov 22:00
0dd8ada

Rebuilt on the design from Anthropic's How we built Claude Code auto mode.

  • Every permission ask goes to the classifier, and so do allowed shell commands that can run arbitrary code
  • The classifier sees only user messages and tool calls, never agent prose or tool output
  • Scripts the agent wrote and then runs are judged by their contents
  • Subagents are judged against the real user's messages
  • Prompt injection probe on tool output
  • Hard and clearable block rules; extend with trustedEnvironment, blockRules, allowRules
  • Default reviewer: DeepSeek V4.1 Flash at high reasoning; optional fast stage 1 via fastModel
  • Any provider model: provider/model#variant
  • 3 blocks in a row tells the agent to ask you; 20 pauses auto mode until your next message

Install: opencode plugin add github:anargia-pixels/opencode-automode-plugin#release

v0.1.0

Choose a tag to compare

@Doctorthe113 Doctorthe113 released this 15 Nov 00:24
4848d58

First release. Install: opencode plugin add @anargia-pixels/opencode-automode. See the README for configuration.