Agent Skills › CopilotKit/OpenTag › fix-tests

fix-tests

GitHub

自动修复代码库或PR中失败的测试。通过定位测试命令、分析失败原因并最小化修改代码，循环执行直至测试通过，最后发布更改或返回日志。

agent/coding/skills/fix-tests/SKILL.md CopilotKit/OpenTag

Trigger Scenarios

用户请求修复失败的测试触发场景为 fix-tests

Install

npx skills add CopilotKit/OpenTag --skill fix-tests -g -y

More Options

Non-standard path

npx skills add https://github.com/CopilotKit/OpenTag/tree/main/agent/coding/skills/fix-tests -g -y

Use without installing

npx skills use CopilotKit/OpenTag@fix-tests

指定 Agent (Claude Code)

npx skills add CopilotKit/OpenTag --skill fix-tests -a claude-code -g -y

安装 repo 全部 skill

npx skills add CopilotKit/OpenTag --all -g -y

预览 repo 内 skill

npx skills add CopilotKit/OpenTag --list

SKILL.md

Frontmatter

{
    "name": "fix-tests",
    "description": "Fix failing tests in a GitHub repo or PR. Use when the brief skill is fix-tests or the user wants the named test command to exit 0."
}

fix-tests

Fix failing tests in the repo from the brief.

Call prepare_repository, including pr_number when kind is pr.
Work on the returned branch; otherwise request opentag/fix-tests-<short-id>.
Find the test command: the brief, then package.json test/ci, then pytest or go test ./... if those files exist.
Run the command. Read the failure. Edit the smallest set of files that fixes it.
Re-run until exit 0 or you cannot fix it.
If green, commit locally and call publish_changes.
If red, do not call publish_changes. Return the log tail.