用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/dust-tt/dust --skill dust-test命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | dust-test |
| description | Step-by-step guide for writing focused, practical tests for Dust codebases following the 80/20 principle. |
Write focused, practical tests for the current file following the 80/20 principle.
When writing tests for a file:
front/tests/utils/factoriesfront/tests/utils/utilsimport {describe, it, expect} from "vitest";
import {makeTestWorkspace, makeTestUser} from "tests/utils/factories";
describe ("ComponentName or FunctionName", () => {
it ("should handle the main happy path", async () => {
// Arrange: Set up using factories
const {workspace} = createResourceTest ()
// Act: Execute the code
const result = await functionUnderTest (workspace);
// Assert: Verify behavior
expect (result).toBeDefined ();
});
it ("should handle the most common edge case", async () => {
// Test the second most important scenario
});
});
Follow similar principles:
describe ("createConversation", () => {
it ("creates conversation with valid params", async () => {
const {workspace, user} = createResourceTest ()
const conversation = await createConversation ({
workspace,
userId: user.id,
title: "Test"
});
expect (conversation.sId).toBeDefined ();
expect (conversation.title).toBe ("Test");
});
it ("fails without required permissions", async () => {
const {workspace, user} = createResourceTest ()
await expect (
createConversation ({workspace, userId: user.id})
).rejects.toThrow ("Permission denied");
});
});
file.test.ts)npm test -- filetotest to verify they passStep-by-step guide for creating Temporal workflows in Dust. Use when adding background jobs, async processing, durable workflows, or task queues.
Review a pull request, branch, commit, diff, or pending changes against Dust's strictest review bar. Use when the user asks to "dust-flavify", "flavify", run a Flavify review, or audit a change against Dust's review standards. Produce a read-only, evidence-backed review that prioritizes existing-pattern reuse, ownership boundaries, database access paths, simplicity, current Dust coding rules, and relevant platform risks. Never post review output or mutate anything on GitHub.
Step-by-step guide for adding support for a new LLM in Dust. Use when adding a new model, or updating a previous one.
基于 SOC 职业分类