跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
搜索与研究

research-routes-run

Execute an approved Research Routes context across software, research-exploration, or hybrid journeys with concurrent route owners, shared control-plane resources, CounterPilot read-only boundaries, explicit evidence and provenance, lifecycle recovery, and HumanDecision seams. Use when implementation or experiments must proceed without treating AI inference as fact or silently changing canonical state.

zenx0x · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add zenx0x/allinluna --skill research-routes-run

技能说明


name: research-routes-run description: Execute an approved Research Routes context across software, research-exploration, or hybrid journeys with concurrent route owners, shared control-plane resources, CounterPilot read-only boundaries, explicit evidence and provenance, lifecycle recovery, and HumanDecision seams. Use when implementation or experiments must proceed without treating AI inference as fact or silently changing canonical state.

Research Routes Run

Execute the approved route graph through the shared control plane and preserve route identity.

  • Load the context and validate it before dispatch.
  • Keep independent route owners concurrent where dependencies permit; share only the shared backbone and declared resources.
  • Run CounterPilot as read-only challenge work. Record its boundary and findings; never let it mutate nodes or promote a route.
  • Require provenance for evidence and preserve all polarity, including failure, conflict, boundary, and unknown.
  • Treat implementation as distinct from canonical. A HumanDecision is required before canonical promotion or setting current continuation.
  • Use lifecycle events to rewind or recover; never rewrite or delete prior route history.
  • Reject missing, cross-context, or ambiguous IDs rather than defaulting to GraphPE, a repository, or a preferred route.
  • Route authorization is a Pack record bound to a confirmed HumanDecision; route authorization alone does not authorize implementation or canonical promotion.
  • A canonical downgrade is an explicit append-only transition that keeps the previous canonical reference and its failure evidence.

Use the shared control-plane, resource, recovery, and router contracts under shared/ together with the existing All in Luna execution machinery. This overlay changes the semantic vocabulary and entry points, not the shared orchestration behavior.