CtrlK
BlogDocsLog inGet started
Tessl Logo

improve-codebase-architecture

扫描代码库中的深化机会,生成可视化 HTML 报告,然后围绕你选中的候选项继续追问。

57

Quality

66%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/engineering/improve-codebase-architecture/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

78%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.

The body is an efficient, highly actionable three-phase workflow with explicit gating between stages and copy-paste-ready commands and phrasings. Its main defect is that its single bundle reference, HTML-REPORT.md, is missing from the bundle, breaking both the promised navigation and the concrete report-building material. Its only other weakness is the absence of validation checkpoints for the report-write/open step.

Suggestions

Add the missing HTML-REPORT.md to the bundle (containing the HTML scaffold, diagram patterns, and styling guidance) so the referenced path resolves and the report step has concrete material

Add a validation checkpoint after writing the temp HTML file — e.g. confirm the file exists and is non-empty before opening it, with a fallback if the open command fails

Tighten the Explore phase with a brief checklist of what the sub-agent must return (candidate list with files/problem/evidence) so the report phase has a deterministic input contract

DimensionReasoningScore

Conciseness

Every line is directive with at most one short sentence of rationale ("深化 module 的收益在于让未来修改更容易"); the body assumes Claude knows git, temp dirs, Tailwind, and Mermaid and explains none of them. No padding or concept explanations are present, matching the lean anchor.

5 / 5

Actionability

Guidance is concrete throughout — `git log --oneline`, $TMPDIR with /tmp and %TEMP% fallbacks, the exact output path `<tmpdir>/architecture-review-<timestamp>.html`, per-OS open commands, exact user-facing quotes, and enumerated card fields. It falls short of 5 because the promised HTML scaffold and diagram patterns are delegated to HTML-REPORT.md, which is not in the bundle, leaving the report-building step without its concrete material.

4 / 5

Workflow Clarity

The three phases (Explore → Present report → Grilling loop) are clearly sequenced with explicit gates ("写完文件后问用户", "用户选中 candidate 后", "现在不要提出 interfaces"). Not 5 because there are no validation steps for the temp-file write or for confirming the report opened successfully, and no error-recovery loop.

4 / 5

Progressive Disclosure

The body itself is well-sectioned and points one level deep to a clearly signaled [HTML-REPORT.md](HTML-REPORT.md), but that file does not exist in the bundle — the only referenced path is broken. Scoring against the actual bundle structure, the navigation promise fails, which sits between the broken/inline structure of 2 and the minor-gap organization of 4.

3 / 5

Total

16

/

20

Passed

Description

53%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.

The description communicates a clear, multi-step capability but omits any "use when" trigger guidance and mixes in second-person phrasing. Trigger-term coverage is narrow, missing the natural architecture/refactoring vocabulary users would reach for. It is distinguishable from neighbors but not sharply so.

Suggestions

Add an explicit "Use when..." clause naming natural triggers, e.g. 'Use when the user asks to improve codebase architecture, find refactoring opportunities, or wants an architecture review'

Rewrite in third person: replace "你选中的候选项" with e.g. "用户选中的候选项" so the description reads as a capability statement

Include the natural keywords users would actually say — architecture, 架构, refactor, 重构, module, architecture review — to improve trigger-term coverage

DimensionReasoningScore

Specificity

It names the domain and three concrete actions ("扫描代码库中的深化机会", "生成可视化 HTML 报告", "继续追问"), which fits the score-4 anchor, but "围绕你选中的候选项" is second person — the third-person rule reduces specificity by 1. It is not 2 because the actions are concrete and domain-specific rather than generic.

3 / 5

Completeness

The "what" is clear (scan for deepening opportunities, generate a visual HTML report, then follow up on selected candidates), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the judging guidelines.

3 / 5

Trigger Term Quality

Relevant keywords exist ("代码库", "深化机会", "HTML 报告") but the natural phrases users would say — "架构", "重构", "refactor", "module" — and their synonyms are entirely missing. It does not reach 4 because keyword coverage lacks any common variations.

3 / 5

Distinctiveness Conflict Risk

The combination of deepening scan + HTML report + follow-up grilling loop is mostly distinct, with only minor overlap risk against generic code-review or refactoring skills. Not 5 because it offers no distinct trigger phrase to disambiguate from those closely related skills.

4 / 5

Total

13

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 1 missing

Warning

Total

15

/

16

Passed

Repository
vinvcn/mattpocock-skills-zh-CN
Reviewed

Table of Contents

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.