CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

huashu-research

结构化网络调研流程,确保调研成果增量保存到文件,不因会话截断丢失。当用户说"调研"、"搜索资料"、"帮我查一下"、"了解一下"、"最新信息"时使用此技能。

Quality

72%

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 ./huashu-research/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

63%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 delivers a clear, actionable research workflow with concrete file conventions and reusable templates, and the incremental-persistence principle is reinforced well. Weaknesses are the dangling SHARED-RULES.md reference that hides a key rule, plus a promotional footer and update date that waste tokens.

Suggestions

Inline the source-priority rules (or ship SHARED-RULES.md in the skill bundle) instead of pointing to an unresolvable file — currently a core rule (信息源优先级) is inaccessible.

Remove the promotional footer ("花叔出品 | 公众号…") and the "最后更新" date line; they consume context without improving task execution.

Add an explicit verification step to the workflow, e.g. after Step 2 confirm each appended finding has a source URL and date before moving on, to strengthen the checkpoint loop.

DimensionReasoningScore

Conciseness

The core workflow and templates are tight, but the promotional footer ("花叔出品 | 公众号「花叔」| 30万+粉丝 … 代表作:小猫补光灯…") and the bare "最后更新: 2026-02-06" date are padding and time-sensitive content that earn no tokens. This matches anchor 3 (mostly efficient, some content to trim) rather than 4.

3 / 5

Actionability

Concrete file path pattern (`_knowledge_base/research-<主题>-<YYYYMMDD>.md`), copy-ready note and briefing templates, and an explicit cadence ("每完成3次搜索…保存一次「阶段摘要」") give mostly executable guidance. It misses anchor 5 because the source-priority rule is delegated to SHARED-RULES.md, which is not present in the skill bundle, leaving a key detail unverifiable.

4 / 5

Workflow Clarity

A clear 4-step sequence (create file → search and append → phase summaries → final briefing) with a periodic checkpoint and a "待确认问题" verification section. It matches anchor 4 rather than 5 because there is no explicit error-recovery or review loop, though the task is neither destructive nor batch so no cap applies.

4 / 5

Progressive Disclosure

Sections are well organized in a single reasonably short file, but the only external reference ("遵循信息源优先级(见 SHARED-RULES.md)") does not resolve within the skill — no references/ or bundle file exists. This matches anchor 3 (references present but not clearly verifiable) rather than 4.

3 / 5

Total

14

/

20

Passed

Description

66%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 has an explicit, well-formed use-when clause with natural Chinese trigger phrases, which is its strongest asset. The main weakness is a high-level "what" that names the domain and persistence mechanism without enumerating concrete actions, plus overly broad triggers that raise conflict risk with general information-lookup requests.

Suggestions

Enumerate the concrete actions in the description, e.g. "创建调研笔记文件、每次搜索后增量追加发现、生成阶段摘要、输出带来源与可信度的最终简报", to raise specificity and completeness.

Narrow the broadest triggers ("了解一下", "帮我查一下") to research-oriented contexts, e.g. "帮我查一下网上的资料" or add qualifiers like "查资料", to reduce conflict with non-research lookups.

State the concrete output artifact in the what-clause ("生成 _knowledge_base/research-<主题>-<日期>.md 调研简报") so users know what they get.

DimensionReasoningScore

Specificity

The description names the domain ("结构化网络调研流程") and 1-2 concrete mechanisms ("增量保存到文件", "不因会话截断丢失"), but does not enumerate the actual actions performed (create research file, append findings, phase summaries, final briefing). It sits at anchor 3 rather than 4 because several specific actions are missing.

3 / 5

Completeness

Both parts are present: an explicit when-clause ("当用户说…时使用此技能") with concrete trigger phrases, and a what-clause. It matches anchor 4 rather than 5 because the "what" ("结构化网络调研流程") is high-level and could list the concrete deliverables more specifically.

4 / 5

Trigger Term Quality

Five natural user phrases ("调研", "搜索资料", "帮我查一下", "了解一下", "最新信息") give good keyword coverage with meaningful variation. It falls short of anchor 5 because common phrasings such as "查资料" or "搜索一下" are not covered.

4 / 5

Distinctiveness Conflict Risk

The persistence-focused research niche is fairly distinct from typical skills, but triggers like "了解一下" and "帮我查一下" are broad enough to fire on non-research information requests. This matches anchor 3 (somewhat specific, still overlapping with similar skills) rather than 4.

3 / 5

Total

14

/

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

alchaincyf/huashu-skills

Path

huashu-research/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.