AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
翻译准备中
A local orchestrator for semi-autonomous agentic SDLC
翻译准备中
取自仓库 README
正在加载完整 README。 以下为已保存的摘录。
A local orchestrator for semi-autonomous software work. One FastAPI process plus a
React SPA: work items enter as chains
— ordered nodes materialized from a YAML template — and each node runs hook-point
tasks through plugin adapters (a headless agent, a subprocess, a builtin). Every
retry loop is capped; hitting a cap escalates to you with the full trace. Gates
stop the chain where a human decision belongs.
Runs on your machine, binds loopback by default, and edits your repos through git
worktrees. How it fits together: ARCHITECTURE.md.查看完整 README ↗目前无法加载完整 README。请通过原文链接阅读。 以下为已保存的摘录。
A local orchestrator for semi-autonomous software work. One FastAPI process plus a
React SPA: work items enter as chains
— ordered nodes materialized from a YAML template — and each node runs hook-point
tasks through plugin adapters (a headless agent, a subprocess, a builtin). Every
retry loop is capped; hitting a cap escalates to you with the full trace. Gates
stop the chain where a human decision belongs.
Runs on your machine, binds loopback by default, and edits your repos through git
worktrees. How it fits together: ARCHITECTURE.md.查看完整 README ↗