searxng-search

Enhanced web and package repository search using local SearXNG instance

3.80x

Quality

63%

Does it follow best practices?

Impact

95%

3.80x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Fix and improve this skill with Tessl

tessl review fix ./searxng-search/SKILL.md

Quality

Content

77%

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 with executable curl/jq examples and a clear start-to-search workflow including readiness checks and debugging guidance. Its main weaknesses are redundant podman/docker duplication and bundle files (two references and one script) that are never linked from the SKILL.md overview.

Suggestions

Link the orphaned bundle files from the body, e.g. add 'See [agent-usage.md](references/agent-usage.md) and [category-guide.md](references/category-guide.md)' and reference scripts/searx where the Nushell helper is discussed.

Reduce redundancy by picking one runtime (or showing the docker variant once) instead of repeating 'podman ... or: docker ...' for every command.

Trim or move the full manual-container-start section behind the start-searxng reference since the script already automates it.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient concrete commands, but dual podman/docker listings for every command and a full manual-container section duplicate the start-searxng script, so it could be tightened; not 3 due to this redundancy, not 1 because it does not pad with concepts Claude already knows.	2 / 3
Actionability	Throughout the body are fully executable, copy-paste-ready curl commands with jq pipes and a quick-reference table (e.g. 'curl "http://localhost:8888/search?q=<query>&format=json"'), matching the score-3 anchor; not 2 because guidance is concrete and complete rather than pseudocode.	3 / 3
Workflow Clarity	For this simple single-task skill the start->search flow is unambiguous, the script 'Wait until ready' is an explicit readiness checkpoint, and the Debugging section gives error-recovery commands; not 2 because validation and error feedback are present, and the destructive/batch cap does not apply to read-only search.	3 / 3
Progressive Disclosure	The two referenced files (package-engine-status.md, pypi-direct-search.md) are well-signaled one-level-deep links, but bundle files agent-usage.md, category-guide.md, and scripts/searx are present yet never linked from the body, so navigation is incomplete; not 3 because orphaned bundle files break easy navigation, not 1 because used references are one-level-deep rather than a monolithic wall.	2 / 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 states a clear domain and mechanism but is a single compound action with no 'Use when' trigger guidance and no natural variation terms for the package ecosystems it targets. It is specific enough to avoid broad conflicts yet too thin to fully satisfy completeness or trigger-term coverage.

Suggestions

Add an explicit trigger clause, e.g. 'Use when searching the web or looking up packages (npm, crates, PyPI) via a local SearXNG instance.'

Replace 'package repository' jargon with natural ecosystem terms users actually say (npm, crates.io, PyPI, packages).

Enumerate concrete actions to lift specificity, e.g. 'Search the web and package repositories (npm, crates, PyPI), filter by category, and return clean JSON.'

Dimension	Reasoning	Score
Specificity	Quotes 'web and package repository search' and 'local SearXNG instance' name the domain and a mechanism, but only a single compound action is given rather than multiple concrete actions, matching the score-2 anchor; not 3 because no list of specific actions, not 1 because a domain and action are named.	2 / 3
Completeness	The 'what' is stated ('Enhanced web and package repository search') but there is no 'Use when...' trigger clause, so per the guidelines completeness is capped at 2; not 1 because 'what' is present, not 3 because 'when' is missing.	2 / 3
Trigger Term Quality	'web... search' and 'package repository search' are relevant keywords, but 'package repository' and 'SearXNG' are technical jargon and common variations (npm, crates, PyPI) are absent, matching score 2; not 3 due to missing natural variations, not 1 because some relevant natural terms are present.	2 / 3
Distinctiveness Conflict Risk	Naming 'local SearXNG instance' gives a specific niche, but 'web and package repository search' is broad enough to overlap generic web-search skills and there are no explicit distinguishing triggers, matching score 2; not 3 because distinct triggers are absent, not 1 because the SearXNG-specific niche limits conflict.	2 / 3
	Total	8 / 12 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: YPares/agent-skills
Commit: ac633fa

Reviewed: 5 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.