CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

lark-doc

飞书云文档(Docx / Wiki)内容操作:读取、创建、编辑文档,插入或下载图片附件,以及操作思维笔记。用户提供文档 URL/token(包括 doubao.com 的 /docx/、/wiki/)时使用;按 URL 路径/token 而非域名路由;`/page/<token>` 是妙搭应用,非飞书云文档。文档内嵌资源按读取参考中的统一规则分流。独立评论操作走 lark-drive;随正文读取评论使用 docs +fetch。表格或 Base 内部数据操作不在本 skill。

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

90%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.

An exemplary routing-index skill body: scenario-first routing to real, well-signaled reference files, read-once discipline, deterministic local-file resolution rules, and an auth error-recovery loop, all with zero padding. The two 4-level dimensions reflect validation being delegated to references for destructive ops, and a partially second-level reference graph plus non-resolvable cross-skill links.

Suggestions

Add a one-line validation/reversal checkpoint next to the destructive shortcuts (+resource-delete, +history-revert, +resource-update) in the routing table so the safety cue exists at SKILL.md level, not only inside the reference files.

Surface the second-level references (genres/ templates, lark-doc-xml.md / lark-doc-xml-extended-blocks.md / lark-doc-md.md) with a brief mention in the relevant routing entries (e.g. under 创建工作流 and +fetch/+update) so they are discoverable from the overview in one hop.

DimensionReasoningScore

Conciseness

The body is a lean ~40-line routing index: every line is a decision rule or scenario→shortcut→reference mapping with zero concept explanations and no padding. Detail is delegated to reference files, fully respecting Claude's competence.

5 / 5

Actionability

Every scenario maps to a named shortcut (+fetch, +update, +media-insert, …) and a real reference file, with concrete flag guidance ('文档操作推荐显式指定 --as user') and deterministic file-resolution rules. As an instruction-only router the guidance is fully specific and executable; absence of inline code is appropriate, not a gap.

5 / 5

Workflow Clarity

The entry sequence is explicit and sequenced ('先判断场景,再读取该场景的参考文件;每个文件只在首次进入对应阶段时读取一次') and includes a feedback loop for auth/scope errors (read lark-shared, fix, retry), plus the whiteboard token-reuse guard. Not a 5 because validation for destructive/batch operations it routes to (+resource-delete, +history-revert) is entirely delegated to references with no checkpoint at this level.

4 / 5

Progressive Disclosure

Clear overview where all 12 referenced files in references/ exist and are one hop from SKILL.md, each clearly signaled by scenario. Not a 5: some bundle content (genres/*.md, lark-doc-xml.md, lark-doc-md.md, lark-doc-xml-extended-blocks.md) is only reachable via a second hop through other references, and the ../lark-whiteboard, ../lark-shared, ../lark-drive links do not resolve within this bundle.

4 / 5

Total

18

/

20

Passed

Description

95%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 description: concrete capability list, explicit 'use when' trigger with URL/path patterns including a cross-product case (doubao.com), and explicit boundary routing against sibling skills. The only gap is that secondary capabilities bundled in the skill (history, whiteboard, cover operations) go unmentioned.

DimensionReasoningScore

Specificity

Names the domain (飞书云文档 Docx/Wiki) and several concrete actions — 读取、创建、编辑文档,插入或下载图片附件,操作思维笔记. Not a 5 because the skill body also covers history revert, whiteboards, and Docx covers, which are absent from the description, leaving minor coverage gaps.

4 / 5

Completeness

Explicitly answers both what ('读取、创建、编辑文档,插入或下载图片附件,以及操作思维笔记') and when ('用户提供文档 URL/token…时使用') with concrete trigger phrases, and adds boundary guidance on what routes elsewhere.

5 / 5

Trigger Term Quality

Covers the natural trigger vocabulary users would actually supply: 文档 URL/token, /docx/ and /wiki/ path patterns (including on doubao.com), 图片附件, 思维笔记, 评论, plus the /page/<token> disambiguation. This is comprehensive natural-term coverage with path patterns and synonyms, matching the top anchor.

5 / 5

Distinctiveness Conflict Risk

Clear niche with explicit conflict routing: 独立评论操作 → lark-drive, 随正文读取评论 → docs +fetch, 表格/Base 数据 → out of scope, /page/<token> → 妙搭应用 not this skill. Minimal risk of triggering the wrong skill.

5 / 5

Total

19

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

relative_links

Relative link issues: 4 suspicious

Warning

Total

13

/

16

Passed

Repository

larksuite/cli

Path

skills/lark-doc/SKILL.md

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.