rudder-docs — Skill Description Optimization

Optimizing your skill's description. This page updates automatically as Claude tests different versions of your skill's description. Each row is an iteration — a new description attempt. The columns show test queries: green checkmarks mean the skill triggered correctly (or correctly didn't trigger), red crosses mean it got it wrong. The "Train" score shows performance on queries used to improve the description; the "Test" score shows performance on held-out queries the optimizer hasn't seen. When it's done, Claude will apply the best-performing description to your skill.

Original: Use when the user asks how Rudder works; needs current Rudder docs, exact behavior, commands, contracts, APIs, or source evidence; asks to create, hire, or configure Rudder agents; or asks to scaffold, develop, or verify Rudder plugins. For explicit creation requests, verify the current interface before acting within existing authorization. Do not use for greetings, ordinary work merely running inside Rudder, routine actions already clear from the active context and typed tools, non-Rudder plugin work, or general coding and research tasks that do not ask about Rudder.

Best: Use when the user asks how Rudder works; needs current Rudder docs, exact behavior, commands, contracts, APIs, or source evidence; asks to create, hire, or configure Rudder agents; or asks to scaffold, develop, or verify Rudder plugins. For explicit creation requests, verify the current interface before acting within existing authorization. Do not use for greetings, ordinary work merely running inside Rudder, routine actions already clear from the active context and typed tools, non-Rudder plugin work, or general coding and research tasks that do not ask about Rudder.

Best Score: 8/8 (test)

Iterations: 1 | Train: 12 | Test: 8

Query columns: Should trigger Should NOT trigger Train Test
Iter Train Test Description Rudder Plugin 的 worker capability、same-origin UI、routePath 和 ctx.assets 现在到底有哪些边界?请查当前 authoring guide、SDK 和源码后回答。 Scaffold a new Rudder Plugin named @acme/quality-gate in a temporary external package directory. Verify the current create-rudder-plugin and SDK workflow first, then run its typecheck, tests, and build. 我需要给 agent 添加 agent-browser。Rudder 的 skills enable 和 skills sync 到底哪个是增量、哪个会替换完整集合?请核对精确策略后回答。 请在 Rudder 里创建一个 Researcher Agent,让它向 CTO 汇报;先核对当前权限、runtime config、role enum 和 approval 流程,再执行创建。 Rudder 里 Chat 和 issue 的职责边界到底是什么?我们团队正在设计交接流程,请给我当前官方文档链接,并说明什么时候应该保留在 Chat。 Before I write this integration, verify which Rudder MCP tools this agent can actually call in the current run. Please use live capability evidence instead of memory. 这个 React 组件在当前 Rudder run 里渲染成空白了,帮我修复状态更新逻辑并补一个组件测试;这不是 Rudder 产品行为问题。 Please fix spelling mistakes on an unrelated library's docs page in this repository, preserve its tone, and run the markdown linter; no Rudder behavior needs clarification. Write a short email to the design team saying the issue is complete and that the new screenshots are ready for review. hi 在这个 Rudder run 里分析附件中的三家竞品定价表,算出年付折扣并推荐我们的套餐价格。 帮我设计一个名叫 Rudder 的船舵图标:深蓝色、圆形徽章、透明背景,输出 1024px PNG。 Before hiring a Founding Engineer in Rudder, compare the installed Codex runtime configuration with existing agents and explain the correct role, title, reportsTo, SOUL.md, desiredSkills, and sourceIssueId fields. How do I configure Codex as a Rudder runtime on the latest released version? I need the supported fields and exact official setup page, not an example from the development branch. 我 checkout 一个已分配的 Rudder issue 时返回了 409。为什么不能直接重试?请对照当前产品契约、源码和测试解释冲突语义。 RUD-248 已分配给我,但标记 done 失败了。请先查清当前 Rudder 对完成评论和验证证据的精确要求,再告诉我应该用哪个受支持的操作。 Use the exposed Rudder skill tool to enable agent-browser for this agent; the exact selection and target agent are already present in the active context. 给这个普通 Vite 应用写一个 markdown-it plugin,把内部链接加上 target 属性;它不是 Rudder Plugin,也不需要 Rudder SDK。 RUD-248 这个 issue 已经分配给我,检查结果也都在上下文里;直接用当前 typed tool 把它标成 done。 The target Agent and new title are already in context. Use the exposed typed Agent update tool to change only its title to Staff Engineer; do not look up Rudder docs.
1 36/36 24/24 Use when the user asks how Rudder works; needs current Rudder docs, exact behavior, commands, contracts, APIs, or source evidence; asks to create, hire, or configure Rudder agents; or asks to scaffold, develop, or verify Rudder plugins. For explicit creation requests, verify the current interface before acting within existing authorization. Do not use for greetings, ordinary work merely running inside Rudder, routine actions already clear from the active context and typed tools, non-Rudder plugin work, or general coding and research tasks that do not ask about Rudder. 3/3 3/3 3/3 3/3 3/3 3/3 0/3 0/3 0/3 0/3 0/3 0/3 3/3 3/3 3/3 3/3 0/3 0/3 0/3 0/3