search-first

Research-before-coding workflow. Search for existing tools, libraries, and patterns before writing custom code. Invokes the researcher agent.

Quality

64%

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 ./.claude/skills/search-first/SKILL.md

Quality

Content

67%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 well structured with a clear workflow, decision matrix, and concrete examples that make the skill actionable. Main weaknesses are token efficiency - the tool-name catalog restates knowledge Claude already has and the ASCII diagram duplicates other sections - and the absence of an explicit validation checkpoint in the workflow.

Suggestions

Trim or remove the "Search Shortcuts by Category" section - the tool names (eslint, prettier, jest, etc.) are knowledge Claude already has and spend tokens without adding guidance.

Collapse the ASCII workflow diagram or merge it with the Decision Matrix, since the five stages and the adopt/extend/compose/build outcomes are described twice.

Add an explicit validation checkpoint after the DECIDE step (e.g. verify the chosen package/MCP installs and works before writing code) to strengthen the workflow's feedback loop.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and assumes Claude's competence, but the "Search Shortcuts by Category" section enumerates well-known tools (eslint, prettier, jest, gofmt...) that Claude already knows, and the ASCII workflow diagram partially duplicates the Decision Matrix and How-to-Use sections. Several sections could be tightened, fitting the score-3 anchor rather than the score-4 'minor trims' anchor.	3 / 5
Actionability	Provides mostly executable guidance: a concrete Task(subagent_type="general-purpose", ...) launch template, a specific "rg" first step, real npm/PyPI search queries, and worked examples with adopt/extend outcomes. Minor gaps remain - the Task snippet uses bracketed placeholders like [DESCRIPTION] and [LANG] rather than fully concrete invocations.	4 / 5
Workflow Clarity	The 5-step workflow (NEED ANALYSIS → PARALLEL SEARCH → EVALUATE → DECIDE → IMPLEMENT) is clearly sequenced and reinforced by a Decision Matrix and a numbered Quick Mode checklist. Minor validation gap: there is no explicit checkpoint confirming the chosen candidate works (e.g. verify the installed package) before implementing, so it sits at 4 rather than 5. This is not a destructive or batch operation, so the workflow-clarity cap does not apply.	4 / 5
Progressive Disclosure	No bundle files exist (no references/, scripts/, or assets/ directories), so the entire skill is a single well-organized file with clearly signaled sections (Trigger, Workflow, Decision Matrix, How to Use, Examples, Anti-Patterns). The ~160-line body inlines a shortcuts catalog and an examples set that could plausibly live in separate reference files, which keeps it at 4 rather than 5.	4 / 5
	Total	15 / 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 clearly states what the skill does with concrete phrasing and a well-defined niche, but it lacks an explicit "Use when..." trigger clause, which weakens completeness and leaves "when to use it" only implied. Trigger term coverage is good though missing common synonyms like "reuse" or "search first".

Suggestions

Add an explicit trigger clause, e.g. "Use when starting a new feature, adding a dependency or integration, or before writing any new utility or helper - or when the user asks to add functionality that might already exist."

Broaden trigger term coverage with natural synonyms users would say, such as "reuse existing code", "don't reinvent the wheel", or "search first".

Mention the concrete search domains (package registries, MCP servers, GitHub) to make the stated actions more comprehensive.

Dimension	Reasoning	Score
Specificity	Names the domain ("Research-before-coding workflow") with 1-2 concrete actions ("Search for existing tools, libraries, and patterns before writing custom code", "Invokes the researcher agent") but coverage is not comprehensive - no search sources, evaluation criteria, or adopt/extend/build outcomes are mentioned. It lists a couple of specific actions rather than several, fitting the score-3 anchor better than the score-4 anchor.	3 / 5
Completeness	Has a clear "what" (search for existing tools, libraries, and patterns) but the "when" is only weakly implied by "before writing custom code" - there is no explicit "Use when..." clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Natural phrases like "Search for existing tools, libraries, and patterns before writing custom code" and "Research-before-coding" are terms a user would plausibly say, giving good keyword coverage. Common synonyms and variations ("reuse", "search first", "don't reinvent the wheel") are missing, so it falls between the 3 and 4 anchors - noticeably above the midpoint.	4 / 5
Distinctiveness Conflict Risk	The niche ("Research-before-coding workflow", searching for existing tools before writing custom code) is clearly distinguishable with distinct triggers. Minor overlap risk remains with general research/search-oriented skills since terms like "research" and "search" are broad, so it fits the score-4 anchor rather than the score-5 anchor.	4 / 5
	Total	14 / 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: devrev/meerkat
Path: .claude/skills/search-first/SKILL.md
Commit: ab88ae9

Reviewed: 8 days 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.