weixin-mp

通过 Limb 体系操作微信公众号：发布文章、上传图片、管理草稿。 Use when: 需要发布内容到微信公众号、查看草稿、检查公众号连接状态。 Not for: 其他平台的发布、纯文本聊天、非公众号相关操作。 Output: 微信公众号操作结果（发布ID、草稿列表、图片URL等）。

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

82%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 highly actionable — every capability ships as an exact, copy-paste-ready Limb invocation with constraints — and the publish workflow is a clear numbered sequence with a pre-flight check, a read-back verification step, and an explicit publish gate. The main gaps are the slightly editorial '冻结编辑真相源' prose and the absence of an explicit verify-fail→fix→re-verify loop in the main workflow.

Suggestions

Tighten the '冻结编辑真相源' step into concrete, imperative checks (extract and pin: title, subtitle/deck, series name+number, author, digest, cover style) and move the editorial rationale into a short rationale line, trimming token overhead.

Make the read-back at step 6 an explicit feedback loop: '核对 media_id/标题/作者/摘要/封面；若不一致，回到 create_draft/update_draft 修正后再次 list_drafts 核对' to satisfy the score-5 validation-recovery anchor for a destructive/publishing operation.

Consider extracting the 核心能力 tool catalog into a references/TOOLS.md with a one-line pointer from SKILL.md, which would let progressive_disclosure reach the score-5 'appropriately split, one-level-deep' anchor.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes Claude's competence — it does not explain what 微信公众号 or the Limb 体系 are — with almost every line being a concrete tool signature or constraint; the only trimming opportunity is the editorial, philosophical prose of the '冻结编辑真相源' sub-step, which keeps it just below the lean score-5 anchor.	4 / 5
Actionability	Every core capability is given as a copy-paste-ready limb_invoke_tool({ nodeId, command, params }) call with exact arguments and concrete constraints (2 MB / 10 MB limits, temp-dir restriction, supported formats), and the workflow is a numbered sequence of those exact calls — a full score-5 match for executable, common-case-covering guidance.	5 / 5
Workflow Clarity	The 发文流程 is a clear 7-step sequence with explicit checkpoints — a 使用前检查 pre-flight, a step-6 list_drafts read-back verification, and a step-7 gate that publishes only on explicit user authorization — but the read-back step does not spell out an explicit 'if mismatch → fix → re-verify' loop, leaving it just below the score-5 feedback-loop anchor.	4 / 5
Progressive Disclosure	A single self-contained file with no bundle directories, organized into clear well-signaled sections (使用前检查、接口发现、核心能力、发文流程、常见错误、限制) and no nested references; because the skill exceeds the 50-line simple-skill threshold and the 核心能力 tool catalog is API-reference-like, it sits just below the score-5 'appropriately split' anchor rather than at it.	4 / 5
	Total	17 / 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.

The description is third-person, concise, and explicitly answers what the skill does and when to use it, with strong trigger-term coverage (synonyms in Chinese and English) and an explicit Not-for boundary. The only minor gap is that the listed actions are broad categories rather than a fully enumerated concrete action list.

Suggestions

Replace the bundled category '管理草稿' with its concrete constituent actions (e.g. 创建草稿、更新草稿、删除草稿、列出草稿) to push specificity toward a fully enumerated score-5 action list.

Consider adding the file/extension context that disambiguates triggers (e.g. mentioning Markdown input and image formats) so the description mirrors the 'including file extensions' breadth of the top anchor.

Dimension	Reasoning	Score
Specificity	Names the domain (微信公众号) and several concrete actions — 发布文章、上传图片、管理草稿、检查连接状态 — but the actions are somewhat general categories (e.g. 管理草稿 bundles create/update/delete/list) rather than the fully enumerated concrete action list of the score-5 anchor, placing it just below 5.	4 / 5
Completeness	Explicitly answers both 'what' (通过 Limb 体系操作微信公众号：发布文章、上传图片、管理草稿) and 'when' (Use when: 需要发布内容到微信公众号、查看草稿、检查公众号连接状态) with concrete trigger phrases, plus a 'Not for' boundary clause — a clear score-5 match.	5 / 5
Trigger Term Quality	Comprehensive natural-term coverage including synonyms and both Chinese and English variants — 微信、公众号、发文、weixin、wechat、publish article — alongside the explicit 'Use when' trigger scenarios, matching the score-5 anchor's coverage including synonyms.	5 / 5
Distinctiveness Conflict Risk	A clear, narrow niche (微信公众号 via Limb 体系) with an explicit 'Not for' clause excluding other platforms, plain chat, and non-公众号 operations, giving minimal conflict risk per the score-5 anchor.	5 / 5
	Total	19 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: zts212653/clowder-ai
Path: packages/api/src/plugins/weixin-mp/skills/weixin-mp/SKILL.md
Commit: 090626a

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

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.