open-source-teardown

明星开源项目拆解：从宣传/PPT/README 进入源码，验证真实架构、明星特性、算法含量、营销水分、可学习点和不 follow 的 tradeoff。 Use when: operator要求拆解热门 GitHub 项目、竞品 agent/runtime、外部 skill/tool 框架，或问“它到底有什么真本事/我们能学什么”。 Not for: 普通资料搜索（用 deep-research）、社区 issue/PR 运营（用 opensource-ops）、只需要架构头脑风暴（用 collaborative-thinking）。 Output: feature-discussions/YYYY-MM-DD-{project}-deep-dive/ 下的代码证据报告 + 对比结论 + 候选 lesson/skill。 GOTCHA: 不许只看 README 下判断；每个明星特性必须追到代码路径、状态突变点、反馈闭环和算法输入输出。

Quality

79%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./cat-cafe-skills/open-source-teardown/SKILL.md

Quality

Content

71%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-organized, action-dense audit methodology with a clear stepwise workflow and concrete commands. Its main weakness is progressive disclosure: the body leans on three refs/*.md templates and one '(internal reference removed)' marker, but none of those bundle files are present, so the skill is not self-contained as written.

Suggestions

Ship the referenced bundle files (refs/report-template.md, refs/teardown-method.md, refs/user-mind-evaluation.md) or inline the essential parts of each so the skill is self-contained and the markdown links resolve.

Remove or replace the '(internal reference removed)' placeholder on the architecture-map line with a concrete pointer (e.g. an inline ASCII/Mermaid example or a real file path).

Tighten the recurring 'claim 真实 ≠ 足以支持产品决定' framing so it is stated once in Step 2 and referenced rather than re-argued across Steps 2, 2.5, and 5.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes Claude's competence — no basic-concept padding — with every section earning its place via audit-specific methodology; minor repetition of the 'claim 真实 ≠ 足以支持产品决定' theme keeps it at 4 rather than the lean 5.	4 / 5
Actionability	Concrete commands (git ls-files, find . -type d -empty, rg, git rev-parse HEAD, pnpm check:skills), explicit paths, and chain/link diagrams give mostly executable guidance; full copy-paste templates are deferred to referenced files, so it is 4 not 5.	4 / 5
Workflow Clarity	A clear Step 0–6 sequence with explicit checkpoints ('commit 后传球', '必须由非作者猫 review') and a Common Mistakes feedback table; validation is peer-review-based rather than automated, placing it at 4 rather than 5.	4 / 5
Progressive Disclosure	The overview is well-structured and signals one-level-deep references ([refs/report-template.md], [refs/teardown-method.md], [refs/user-mind-evaluation.md]), but none of those files exist in the bundle (no refs/, references/, scripts/, or assets/), so the signaled navigation is broken — matching the 3 anchor where references are present but do not resolve.	3 / 5
	Total	15 / 20 Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description that crisply states the audit purpose, gives explicit Use-when/Not-for/Output guidance, and disambiguates from sibling skills. Specificity and trigger-term coverage are solid but stop just short of fully comprehensive, capping those two dimensions at 4.

Dimension	Reasoning	Score
Specificity	Lists several concrete audit actions — '验证真实架构、明星特性、算法含量、营销水分、可学习点和不 follow 的 tradeoff' and '从宣传/PPT/README 进入源码' — but they are audit categories rather than the ultra-concrete operations of the 5 anchor, so it sits at 4 not 5.	4 / 5
Completeness	It explicitly answers what (source-code-first capability teardown), when ('Use when: operator要求拆解…'), and adds 'Not for' exclusions plus an 'Output' spec — clearly matching the 5 anchor for concrete what + when.	5 / 5
Trigger Term Quality	The 'Use when' clause surfaces natural operator phrases like '拆解热门 GitHub 项目', '它到底有什么真本事', '我们能学什么', and 'marketing vs reality' with synonyms, giving good keyword coverage; a few common variations are absent so it is 4 rather than 5.	4 / 5
Distinctiveness Conflict Risk	An explicit 'Not for' block distinguishes deep-research, opensource-ops, and collaborative-thinking, carving a clear niche (source-code-first capability audit) with minimal conflict risk, matching the 5 anchor.	5 / 5
	Total	18 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 3 missing	Warning

	Total	14 / 16 Passed

Repository: zts212653/clowder-ai
Path: cat-cafe-skills/open-source-teardown/SKILL.md
Commit: 090626a

Reviewed: ~23 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.