菜单
Prepare, run, diagnose, and clean up this repository's end-to-end tests securely. Use when running `example/e2e` or `packages/sdk/e2e`, fixing their failures, or handling authentication, workspace, deployed-state, or cleanup prerequisites.
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。
Use this skill when working with @tailor-platform/sdk projects, including service configuration, CLI usage, and docs navigation.
Run and maintain the llm-challenge evidence collector for SDK affordance work. Use when the user mentions llm-challenge, challenge runs, challenge results, or creating llm-challenge problems.
Search and retrieve Tailor Platform documentation from docs.tailor.tech. Use this skill whenever working with platform services -- even when the codebase seems to have the answer. SDK code shows implementation details, but the official docs describe authoritative platform behavior, configuration options, and constraints that may not be apparent from code alone. Covers: TailorDB, Auth, Executor, Function, Workflow, Resolver, Secret Manager, Static Website, AppShell, TRN format, platform limits, and more. Also use during planning phases when platform capabilities or limitations need clarification.
基于 SOC 职业分类
| name | e2e-test |
| description | Prepare, run, diagnose, and clean up this repository's end-to-end tests securely. Use when running `example/e2e` or `packages/sdk/e2e`, fixing their failures, or handling authentication, workspace, deployed-state, or cleanup prerequisites. |
| metadata | {"internal":true} |
Run the selected suite against authenticated, current platform state and clean up resources created by the run.
example/e2e, packages/sdk/e2e, or both from the user's request. Ask when the request is
open-ended; the suites have different prerequisites.| Suite | Workspace lifecycle | Pre-deploy |
|---|---|---|
example/e2e | Reuses one workspace | Deploy example/ when drifted |
packages/sdk/e2e | Creates disposable workspaces | Test deploys itself |
Read .agents/skills/e2e-test/ids.local.env first. It is gitignored and may contain only:
TAILOR_PLATFORM_WORKSPACE_ID for example/e2eTAILOR_PLATFORM_ORGANIZATION_ID for packages/sdk/e2eTAILOR_PLATFORM_FOLDER_ID for packages/sdk/e2eIf the file is missing but .agents/skills/e2e-setup/ids.local.env exists, move that legacy file to
the new location and set mode 0600. Never store tokens, client IDs, or client secrets in this file.
A valid TAILOR_PLATFORM_PROFILE can supply the example workspace without this file.
Never source the file. Run commands through the loader, which accepts only the three UUID fields and rejects duplicate, malformed, or executable content:
/bin/bash .agents/skills/e2e-test/scripts/with-e2e-ids.sh \
.agents/skills/e2e-test/ids.local.env -- <command> [args...]
packages/sdk/e2e: inspect both the test status and exact-namespace cleanup status