CtrlK
BlogDocsLog inGet started
Tessl Logo

chengfeng-cut-talking-head

剪辑中文口播原素材:逐词转录、按词典修正听错的专名、识别口误与重复、生成删词候选、打开 Studio 让用户复核。只产出一份已复核的删词账本,不切媒体、不做字幕、不做分镜动画。用户说剪口播、处理口误、生成口播基础素材、继续剪口播,或确认卡回传 action=return_cut_review 时使用。不要用于执行物理剪切、导出剪后视频、单独安装、单独打开工作台或口播分镜成片。

72

Quality

88%

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

SKILL.md
Quality
Evals
Security

Quality

Content

77%Weight 40%Scale 1-3

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

The body is highly actionable with executable commands and a well-sequenced, validation-gated workflow, but it is padded with rationale and inline dated war-stories, and two of its three external references are absent from the bundle with no clean navigation block.

Suggestions

Move the dated incident notes (2026-07-26, 2026-07-28) and the "为什么…" rationale blocks into a dedicated 'old patterns / lessons learned' or reference file so the main flow stays lean and the time-sensitive context doesn't inflate the working instructions.

Add a short navigation block listing each external reference with its purpose, and confirm business-workflow-contract.md and runtime-and-product-contract.md are actually shipped in the bundle — otherwise inline the minimum needed contract steps so the skill is self-contained.

Tighten or relocate the 修字 war-story detail (e.g. the 16-写法 / 39-词 anecdote) into semantic-deletion.md so the instructional steps in the main body read as compact directives.

DimensionReasoningScore

Conciseness

It assumes Claude's competence (no basic-concept teaching) and leads with commands, but is padded with rationale blocks ("为什么不切…", "为什么必须在判断之前…") and inline dated incident notes (2026-07-26, 2026-07-28) not isolated in a deprecated section, so it is mostly efficient but could be tightened.

2 / 3

Actionability

Provides fully executable, copy-paste-ready commands with concrete flags (e.g. `node "$VC" project create "$jobDir" --video … --transcript … --json`) plus a complete JSON candidate schema, leaving no guesswork.

3 / 3

Workflow Clarity

Clearly sequenced sections 0–5 with explicit validation checkpoints ("只有脚本确认服务 healthy=true…才继续", readback projectId/revision matching, readBackVerified) and a recovery section ("恢复与失败") giving fix→retry feedback loops for risky batch/destructive operations.

3 / 3

Progressive Disclosure

It has numbered section structure and points to one real reference (references/semantic-deletion.md, present), but two referenced files (../../references/business-workflow-contract.md, ../../references/runtime-and-product-contract.md) are missing from the bundle and references are scattered inline rather than in a clearly signaled navigation block, with detail that could live separately kept inline.

2 / 3

Total

10

/

12

Passed

Description

100%Weight 40%Scale 1-3

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 is specific, trigger-rich, and complete, clearly stating what the skill does, when to invoke it, and what it deliberately does not do. It uses third-person imperative voice and avoids vague fluff.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — "逐词转录、按词典修正听错的专名、识别口误与重复、生成删词候选、打开 Studio 让用户复核" — each a specific operation rather than vague language.

3 / 3

Completeness

Explicitly answers both what (the transcription/修正/候选/复核 actions and the single deliverable "已复核的删词账本") and when ("用户说剪口播、处理口误…时使用"), with an explicit Use-when clause.

3 / 3

Trigger Term Quality

Covers natural trigger phrases users would say — "剪口播、处理口误、生成口播基础素材、继续剪口播" — plus a concrete callback trigger (action=return_cut_review), giving good coverage of likely utterances.

3 / 3

Distinctiveness Conflict Risk

Occupies a clear niche (中文口播删词账本) and explicitly fences off adjacent work — "不切媒体、不做字幕、不做分镜动画" and "不要用于执行物理剪切、导出剪后视频…" — making wrong-skill triggering unlikely.

3 / 3

Total

12

/

12

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.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 2 suspicious

Warning

Total

15

/

16

Passed

Repository
Agentchengfeng/chengfeng-videocut-skills
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.