knowledge

本地知识库集成 - 文档检索、投喂、双轨模式切换

Quality

61%

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

Fix and improve this skill with Tessl

tessl review fix ./configs/microservice/bff-service/configs/agent-skills/clawhub/knowledge/SKILL.md

Quality

Content

72%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is admirably concise and well-organized for a simple skill, scoring top marks on token efficiency and structure. Its weaknesses are actionability and workflow clarity: the commands are natural-language triggers rather than executable calls, and the dual-track mode switch lacks any validation checkpoint.

Suggestions

Make the commands executable: show the exact API call or script for searching the knowledge base (e.g. an HTTP request to http://127.0.0.1:8001) rather than only the trigger phrase.

Add a verification step after a mode switch (e.g. confirm the active mode by querying the service state) so the dual-track switch has a feedback checkpoint.

Clarify the '投喂' action with a concrete example of how a document is added to E:/knowledge-base.

Dimension	Reasoning	Score
Conciseness	The body is a lean command table plus two config lines with no padding or explanation of concepts Claude already knows; every token earns its place. Not below 3 because there is no verbosity or unnecessary explanation.	3 / 3
Actionability	It gives concrete commands ('查一下 xxx', '切换到本地知识库') and concrete paths/endpoints, but the commands are natural-language triggers rather than executable code or exact API calls, leaving key implementation details unspecified. Not a 3 because guidance is not fully executable/copy-paste ready; not a 1 because specific commands and endpoints are present.	2 / 3
Workflow Clarity	The command table maps triggers to actions but there is no sequence, and switching between the two modes (dual-track) has no validation or verification step despite being a mode-changing operation. Per the feedback-loop note, a mode switch without a verification checkpoint caps this at 2; not a 1 because the single actions are unambiguous.	2 / 3
Progressive Disclosure	This is a simple, single-purpose skill well under 50 lines with no bundle files, organized into clear sections ('使用方法', '文件位置'); per the simple-skills note this can score 3. Not below 3 because the structure is clear and there is nothing that should be split out.	3 / 3
	Total	10 / 12 Passed

Description

50%

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 names a clear niche and several concrete actions but is terse and lacks any explicit 'when to use' trigger guidance, capping several dimensions at 2. It uses appropriate third-person voice and is concise, but trigger-term coverage and completeness are the weak points.

Suggestions

Add an explicit trigger clause, e.g. 'Use when the user wants to search a local knowledge base, feed it new documents, or switch between local retrieval and AnythingLLM conversational mode.'

Replace jargon like '投喂' and '双轨模式' with natural user-facing phrasing (e.g. 'add documents to the knowledge base', 'switch between local search and conversation mode') to improve trigger-term quality.

Expand the capability list with one or two more concrete actions (e.g. 'view document counts') to lift specificity toward a 3.

Dimension	Reasoning	Score
Specificity	The phrase '文档检索、投喂、双轨模式切换' names several concrete actions (retrieval, feeding, dual-mode switching) but they are terse and read as functional labels rather than a comprehensive list of specific capabilities; not a 3 because it lacks the rich, multiple concrete actions seen in the top anchor.	2 / 3
Completeness	It states what the skill does (the actions) but provides no 'Use when...' or equivalent explicit trigger guidance, which per the judging guidelines caps completeness at 2; not a 3 because the 'when' is entirely missing rather than explicit.	2 / 3
Trigger Term Quality	Terms like '文档检索' are reasonably natural, but '投喂' and '双轨模式' are domain jargon and there are no natural user-spoken trigger variations; not a 1 because it is not purely technical jargon, but not a 3 because it lacks good coverage of common natural phrasings.	2 / 3
Distinctiveness Conflict Risk	'本地知识库集成' with a dual-track mode is a somewhat specific niche that is unlikely to broadly conflict, but the description is terse enough that it could still overlap with general retrieval/document skills; not a 3 because the triggers are not distinct enough to rule out overlap.	2 / 3
	Total	8 / 12 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/clawhub/knowledge/SKILL.md
Commit: f618458

Reviewed: about 14 hours 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.