CtrlK
BlogDocsLog inGet started
Tessl Logo

pipeline-model-architecture

理解或修改 BK-CI 流水线 Model、Stage、Container、Element 结构时使用。当用户提到流水线模型、Stage/Job/Task 数据结构、模型扩展、模型校验、版本兼容、YAML 与 Model 转换时优先使用。

61

Quality

77%

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 ./ai/skills/pipeline-model-architecture/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

65%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, information-dense architecture overview that succeeds at conciseness and gives genuinely concrete BK-CI-specific guidance and pitfall checklists. It is held back by a broken disclosure layer — every detailed reference points to files missing from the bundle — and by an entry-step mismatch (5 claimed categories vs 3 listed) that blurs the routing workflow.

Suggestions

Fix the routing step: either list all 5 problem categories explicitly or change "归到下面 5 类之一" to match the 3 documents actually offered, so the classification entry point is unambiguous.

Ship the referenced files (reference/1-core-structure.md, reference/2-persistence-versioning.md, reference/3-runtime-conversion-extension.md) in the bundle under references/, or remove the dead pointers until they exist — currently every 'read this next' path dead-ends.

De-duplicate the reference list by keeping it only in 延伸阅读 (or only in 快速指导), and make the cross-skill pointers to process-module-architecture and yaml-pipeline-transfer resolvable as well.

DimensionReasoningScore

Conciseness

Lean and dense — every section (适用场景, 快速指导, 高信号规则, 关键陷阱) delivers BK-CI-specific knowledge Claude would not already have, with no concept padding. Not a 5 only because the three reference documents are listed twice (in 快速指导 item 1 and again in 延伸阅读), a trimmable redundancy.

4 / 5

Actionability

Concrete, specific guidance throughout for an instruction-only skill: exact touch-points when adding a polymorphic type ("Jackson 多态序列化配置", "前端编排模型", "后端转换与校验", "运行引擎或 Worker 识别逻辑") and an explicit verification list ("历史 Model 回放、重试、版本切换和 YAML 转换"). Minor gaps: the routing targets it points to are not present, and the classification scheme is stated as 5 categories but only 3 are listed.

4 / 5

Workflow Clarity

The routing-first step is a sound entry point and 关键陷阱 functions as implicit checkpoints, but the entry instruction is incoherent — "先把问题归到下面 5 类之一" while only 3 categories/documents are listed — and 快速指导 items 2–6 are unsequenced heuristics rather than an ordered workflow.

3 / 5

Progressive Disclosure

The overview/pointing structure is good and references are clearly signaled one level deep, but none of the referenced files (reference/1-core-structure.md, reference/2-persistence-versioning.md, reference/3-runtime-conversion-extension.md) exist in the bundle — no reference/references directory is present — so navigation dead-ends; the path also uses 'reference/' rather than the conventional 'references/'.

3 / 5

Total

14

/

20

Passed

Description

78%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 well-scoped, explicitly triggered description in the BK-CI pipeline-model niche: it answers both what and when with concrete, natural trigger phrases and is highly distinguishable from sibling skills. Its main limitation is that the 'what' relies on two generic verbs (understand/modify) plus a list of structures rather than enumerating concrete actions.

DimensionReasoningScore

Specificity

Names the domain and concrete objects/operations ("Model、Stage、Container、Element 结构", "模型校验", "版本兼容", "YAML 与 Model 转换"), but the verbs are limited to "理解或修改" (understand or modify), which is less comprehensive than anchor 5's multiple concrete actions.

4 / 5

Completeness

Both parts are present: what ("理解或修改 BK-CI 流水线 Model、Stage、Container、Element 结构") and an explicit when-clause ("当用户提到…时优先使用") with concrete triggers. Not a 5 because the 'what' side lists only two generic verbs rather than concrete capabilities.

4 / 5

Trigger Term Quality

Good coverage of natural trigger phrases ("流水线模型", "Stage/Job/Task 数据结构", "模型扩展", "模型校验", "版本兼容", "YAML 与 Model 转换") including the Stage/Job/Task synonyms, though common related phrasings like 插件/Element 新增 or 序列化 are missing.

4 / 5

Distinctiveness Conflict Risk

Clear niche — the BK-CI pipeline model layer specifically — with distinct trigger terms; neighboring concerns (execution chain, YAML/PAC) are explicitly deferred to other skills in the body, so conflict risk is minimal.

5 / 5

Total

17

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
TencentBlueKing/bk-ci
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.