Content
90%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A highly efficient, actionable CLI skill body: executable commands, explicit validation and error-handling rules, and proper offloading of live-stream detail to a real reference file. The only structural blemish is a reference to a `tools/douyin-login` path that does not exist in this bundle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~45-line body is lean and dense with operational facts only (session handling, exit codes, ID rules, pagination caps, delegation boundaries); it explains nothing Claude already knows and every sentence earns its place. Not below 5 because no padding or over-explanation is present. | 5 / 5 |
Actionability | The bash block gives copy-paste-ready commands covering the common cases (search by video/user, fetch with media download, comments and replies, user-posts, collect, methods/call discovery), plus concrete outputs (note.json, video.mp4, image-01.jpg) and explicit pagination/error behavior. | 5 / 5 |
Workflow Clarity | Validation checkpoints are explicit — "核对每条作品的 author.sec_uid...作者不符时停止后续账号分析" and "缺列表字段或游标异常报错，不当作空数据成功" — with a feedback loop (re-read original JSON when only a truncated ID is available). It is not 5 because there is no unified step ordering across the operations; each rule is stated per-command. | 4 / 5 |
Progressive Disclosure | The body is well-sectioned and the live detail is appropriately split into the clearly signaled, one-level-deep [直播读取说明](references/live.md), which exists in the bundle. Minor gap: the body cites `tools/douyin-login`, which is not present in this bundle's file structure. | 4 / 5 |
Total | 18 / 20 Passed |