test: fix ci workflow#38
Conversation
hustcc
commented
Apr 9, 2026
- make ci work
There was a problem hiding this comment.
Code Review
This pull request significantly reduces test coverage by deleting several unit test files and the testing documentation, while adding a new integration test for the retrieval API and migrating the Vitest configuration to TypeScript. The review feedback emphasizes the critical need to restore the deleted unit tests to prevent regressions and recommends updating the testing documentation. Additionally, an improvement was suggested for the new integration tests to include more robust assertions for library filtering.
I am having trouble creating individual review comments. Click here to see my feedback.
tests/bm25.test.js (1-229)
This pull request removes a significant number of test files, including this one for the BM25 logic, as well as tests for the skill builder and validator. While the goal may be to fix the CI workflow, deleting tests drastically reduces coverage for critical components like query cleaning, tokenization, and synonym expansion. This introduces a high risk of regressions going unnoticed in the future.
The new retrieve.test.ts provides some integration-level testing, but it does not replace the fine-grained unit tests that were present in this file.
Please consider restoring these tests. If they were flaky, they should be fixed rather than removed. If the underlying code is no longer used, it should be removed in a separate, clearly-documented change.
tests/README.md (1-59)
This README.md file, which explained the testing strategy, network requirements, and test coverage, has been removed. This documentation was valuable for developers to understand and contribute to the test suite. Please consider adding a new, updated README.md in the __tests__ directory to document the current testing approach.
tests/retrieve.test.ts (19-22)
The assertion expect(Array.isArray(results)).toBe(true) is very weak, as the retrieve function will likely always return an array. A better test would be to assert that any returned results actually belong to the 'g6' library. This would more accurately test the library filtering functionality.
it('should support g6 library parameter', () => {
const results = retrieve('graph layout', 'g6');
expect(Array.isArray(results)).toBe(true);
// If results are returned, they should all be from the 'g6' library.
results.forEach(skill => {
expect(skill.library).toBe('g6');
});
});* chore: 新增 g2 skills * chore: add performance part * fix: test * fix: Ai CR * chore: user invocable * chore: AI SDK 优化 * chore: add harness * fix: path error * chore: 补充参考文档 * chore: 新增 grep 兜底检索 * chore: eval 自检更新 * chore: 删除无效scripts * chore: skill 优化 * chore: add worktree * chore: add kimi model * chore: override env * fix: 修复 data 格式错误 * chore: 目录结构调整 * chore: 补充评测数据 * chore: 删除 common 模块 * chore: 优化 scale skill * chore: skill tool 移动到 cli 下 * chore: add playground * refactor: skill cli (#37) Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> Co-authored-by: Joel Alan <31396322+lxfu1@users.noreply.github.com> * test: fix ci workflow (#38) Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> * docs: add readme for cli usage (#39) * docs: add readme for cli usage * docs: fix skill typo * docs: update * test: fix ci --------- Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> * refactor: playground for cli and skill (#40) * chore: 补全 playground libs * fix: ts * fix: resolve conversation * chore: 代码优化 * fix: 修复由于 nextjs 加载源码导致 __dirname 异常 (#44) * chore: 添加 claude plugin (#43) * refactor: eval logic (#42) * chore: 优化 harness 逻辑 (#48) * chore: 优化 harness 逻辑 * chore: 优化 harness 流程 * fix: result count * feat: 新增 retrieveWithContent (#49) * feat: 新增 retrieveWithContent * fix: test * chore: loadSkillContent 目录约束 * fix: 修复 work tree 模式下 __dirname 不一致导致的优化异常 * chore: 新增 harness 优化 skill 后的基线对比逻辑 (#51) * chore: cli 重命名 (#52) * chore: cli 重命名 * fix: tests * refactor: antv cli (#53) * 精简索引构建输出并优化检索链路(移除 indexDir、使用 gray-matter 解析元数据、简化 includeContent,并统一 index 到 src/index→dist/index) (#56) * Refactor playground AI flow to Vercel AI SDK with Skill/CLI mode tool assembly and token usage visibility (#54) * Initial plan * refactor(playground): migrate chat flow to vercel ai sdk Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/60eb740b-946b-4dbc-8065-b2143da12fa4 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): harden skill path handling and restore temperature Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/60eb740b-946b-4dbc-8065-b2143da12fa4 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): split skill tools into separate modules Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/bcdb3e77-18a5-4b73-8c9d-4ae97f87510a Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * feat(playground): redesign dual-mode ai tools workflow Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/dd7b7fe2-54a4-4e8c-ac4d-531df443ff31 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): normalize library naming in prompts Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/dd7b7fe2-54a4-4e8c-ac4d-531df443ff31 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): tighten tool input limits and constants Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/dd7b7fe2-54a4-4e8c-ac4d-531df443ff31 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): simplify read-file validation fallback naming Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/dd7b7fe2-54a4-4e8c-ac4d-531df443ff31 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): use explicit step constants and schemas Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/dd7b7fe2-54a4-4e8c-ac4d-531df443ff31 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): rename modes to skill/cli and unify tools directory Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/69d26ddf-26a6-43dd-bdca-32bb851ea746 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): harden tools path handling and naming Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/69d26ddf-26a6-43dd-bdca-32bb851ea746 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): tighten tool path validation guards Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/69d26ddf-26a6-43dd-bdca-32bb851ea746 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): move mode entry files above tools directory Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/e4438102-2cd2-42ac-b772-a29fa11c5a20 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * fix(playground): adapt retrieve tool to updated retrieve api Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/8d505619-d6d9-48dd-85a5-af7c42bc73b7 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): remove retrieve content fallback path Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/c8187137-e11b-480a-9abf-ed20f30bb6f7 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * refactor(playground): use gray-matter in list references tool Agent-Logs-Url: https://github.com/antvis/chart-visualization-skills/sessions/23696116-a1cb-402e-954e-06b426ddc4b2 Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> * chore: opt cli command * refactor: playground * feat: fix style * feat: update website meta info * fix: tool call error --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> Co-authored-by: hustcc <i@hust.cc> Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> * refactor: 重构 g2-chart skill (#57) * refactor: 重构 g2-chart skill * chore: resolve skill * chore: 去掉 eval 中 list refercnces 逻辑 * chore: playground 模块支持加载外层 env (#58) * chore: playground 模块支持加载外层 env * Apply suggestions from code review Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> --------- Co-authored-by: hustcc <i@hust.cc> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> * docs: add agent.md (#60) * docs: add agent.md * chore: fix typo --------- Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> * chore: 修改 env 示例 (#59) * chore: 修改 env 示例 * chore: 新增 VL Model 配置示例 * refactor: playground (#61) * refactor: playground * Update playground/src/libs/cli-mode.ts Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> * Update playground/src/libs/skill-mode.ts Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> * Update playground/src/app/api/generate/route.ts Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> --------- Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> --------- Co-authored-by: hustcc <i@hust.cc> Co-authored-by: 逍为 <xiaowei.wzw@antgroup.com> Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com> Co-authored-by: hustcc <7856674+hustcc@users.noreply.github.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>