tavily-search

Web search using Tavily's LLM-optimized API. Returns relevant results with content snippets, scores, and metadata.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

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

Quality

Content

92%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 a strong example of a simple script-backed skill: fully executable commands, accurate option documentation matching the bundled script, and tight tables instead of prose. The only real improvement is removing the duplication between the Quick Start command list and the Examples section.

Dimension	Reasoning	Score
Conciseness	The body is lean — auth setup, runnable commands, two compact tables, and five terse tips — with no explanation of concepts Claude already knows. Not 5 because the 'Quick Start > Using the Script' block and the following 'Examples' section largely repeat the same invocations, which could be merged or trimmed.	4 / 5
Actionability	Every command is copy-paste executable (node {baseDir}/scripts/search.mjs "query" -n 10 / --deep / --topic news / --include-domains), the options table gives flags with defaults, and the examples cover the common cases including domain filtering and result counts. The referenced script (scripts/search.mjs, 150 lines with matching usage text) is real and its documented flags match the table.	5 / 5
Workflow Clarity	This is a simple single-purpose skill (run one search command) and the single action is unambiguous: authenticate via TAVILY_API_KEY, then invoke the script with the documented options. The operation is read-only, so no validation checkpoints are required and the simple-skill exception applies.	5 / 5
Progressive Disclosure	Content is appropriately split: the executable logic lives in the one-level-deep bundle file scripts/search.mjs referenced via clear {baseDir} paths, while the SKILL.md body stays a well-organized overview (Authentication, Quick Start, Options, Search Depth, Tips) with everything else inline and short. Navigation is easy and nothing is buried or nested.	5 / 5
	Total	19 / 20 Passed

Description

61%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 gives a clear, concise picture of what the skill does and what it returns, with the natural trigger term 'web search'. Its main gap is the complete absence of when-to-use guidance, which caps completeness, and a capability list that understates the options documented in the body.

Suggestions

Add an explicit trigger clause, e.g. 'Use when the user asks to search the web, look up current information, or find recent news/online sources.'

Surface one or two more concrete capabilities from the body, such as advanced/deep search mode, news topic filtering, and domain include/exclude filtering, to raise specificity.

Mention time-bounded queries (e.g., 'for recent information within a day/week/month') as a differentiating trigger against general search skills.

Dimension	Reasoning	Score
Specificity	"Web search using Tavily's LLM-optimized API" names the domain, and "Returns relevant results with content snippets, scores, and metadata" describes the output, but the only concrete action is generic web search — no mention of depth modes, topic filtering, or domain filtering. This matches the anchor 'Names domain and 1-2 concrete actions, but not comprehensive'; not 4 because coverage of capabilities is a single action plus an output description.	3 / 5
Completeness	The 'what' is clear (web search via Tavily returning snippets, scores, metadata), but there is no 'Use when...' clause or any trigger guidance — the rubric explicitly caps completeness at 3 in that case. Not 2 because the 'what' half is concrete and specific, not vague.	3 / 5
Trigger Term Quality	"Web search" is exactly the natural phrase a user would say when needing this skill, plus "search the web" equivalents are implied. Not 5 because common variations like "look this up", "find recent info", or "research" are absent.	4 / 5
Distinctiveness Conflict Risk	Naming Tavily gives it a clear niche distinct from general document or code skills, with only minor overlap risk against other web-search skills (e.g., a built-in search tool). Not 5 because without trigger guidance, any request phrased as 'search' could plausibly route here or to a competing search skill.	4 / 5
	Total	14 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/clawhub/tavily-search/SKILL.md
Commit: a16301c

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.