CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

deprecate-workflow-node

当用户需要弃用一个工作流节点(保留向后兼容、隐藏出模板面板)时触发该 skill。FastGPT 工作流节点的弃用流程标准化封装,覆盖模板、Dispatcher、UI 引用等所有需要改动的位置。

66

Quality

83%

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

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.

Excellent actionability and a well-validated workflow for a risky refactor operation, written almost entirely from repo-specific knowledge. The weaknesses are structural: no progressive disclosure into reference files, some triplication of the step list, and an implicit rather than explicit failure-recovery loop.

Suggestions

Move the 详细步骤 sections into a reference file (e.g. references/steps.md) and keep SKILL.md to the TODO checklist, 核心理念 table, and verification checklist — this would cut the always-loaded body substantially.

State the error-recovery loop explicitly after step 8 (e.g. 'if typecheck fails on import paths, re-check the ../ depth in the moved files and rerun') to convert the implicit retry into a feedback loop.

Deduplicate the three restatements of the step list (TODO template, detailed steps, anti-patterns) — for example, fold the anti-pattern column of the 核心理念 table into the 反模式 section.

DimensionReasoningScore

Conciseness

Almost every token carries repo-specific knowledge Claude cannot infer (exact paths, moduleTemplatesFlat semantics, the dead `abandon` field warning, shared-child-node caveats). The score sits below 5 only because the 核心理念 table, TODO template, and 详细步骤 restate the same 8 steps three times at different altitudes.

4 / 5

Actionability

Fully executable throughout: copy-paste grep commands with --include filters, exact source/target paths for git mv, before/after code diffs for constants.ts changes, the SoonOffline enum snippet, and pnpm lint/test commands. The <enumKey> placeholders are justified by the explicit fill-in TODO template.

5 / 5

Workflow Clarity

An 8-step TODO template, matching detailed steps, a final verification checklist (template panel search, old-workflow load/run, badge rendering, lint), and a lint/typecheck/partial-test step give clear checkpoints. It misses 5 only because the error-recovery loop (what to do when lint/typecheck or a checklist item fails) is implicit rather than stated.

4 / 5

Progressive Disclosure

A ~205-line monolithic SKILL.md with no bundle files at all. Sections are clearly organized, but the eight detailed step sections with code examples are exactly the deep material that belongs in a one-level reference file, leaving SKILL.md as the overview plus TODO checklist.

3 / 5

Total

16

/

20

Passed

Description

80%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 that explicitly states both what it does and when to trigger, anchored to a specific repo and mechanism. Its main weakness is thin trigger-term coverage — a single keyword where synonyms and node-type examples would catch more user phrasings.

Suggestions

Add the synonyms from the body ("废弃 / 下线 / 不再推荐使用") and one or two example node names (e.g. LoopNode, RunAppModule) to the description so more natural user phrasings trigger the skill.

Name one or two of the concrete operations (e.g. moving templates/Dispatchers to abandoned/ and adding a SoonOffline status badge) to push specificity from 'coverage areas' to 'concrete actions'.

DimensionReasoningScore

Specificity

The description names the domain (FastGPT workflow node deprecation) and several concrete coverage areas: "保留向后兼容、隐藏出模板面板", "覆盖模板、Dispatcher、UI 引用等所有需要改动的位置". It stops short of enumerating the actual operations (file moves, @deprecated annotations), which is the gap that keeps it below 5.

4 / 5

Completeness

Both halves are explicit: what ("FastGPT 工作流节点的弃用流程标准化封装,覆盖模板、Dispatcher、UI 引用等所有需要改动的位置") and when ("当用户需要弃用一个工作流节点…时触发该 skill").

5 / 5

Trigger Term Quality

It has a natural, explicit trigger phrase ("当用户需要弃用一个工作流节点…时触发该 skill"), but relies on a single keyword ("弃用"); the synonyms users would actually say ("废弃", "下线", "不再推荐") appear only in the body, not the description.

3 / 5

Distinctiveness Conflict Risk

A clear niche with distinct triggers: deprecating FastGPT workflow nodes specifically, including the backward-compatibility and template-panel constraints that separate it from generic refactoring or deletion skills.

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

labring/FastGPT

Path

.agents/skills/core/workflow/deprecate_workflow_node/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.