core-vitals-fixer

SAFE max severity MEDIUMSkillSpector Safe 12 registry clawhub.ai last scanned 2026-02-05 1 scan

Verdict safe, but 2 findings (2 medium), mainly social engineering, data exfiltration. Most recent scan 2026-02-05.

This skill presents MEDIUM security concerns primarily due to misleading representation and undisclosed external data transmission. The skill masquerades as an agent skill but is actually just instructions to run an external NPX package that sends user code to OpenAI's API. The main issues are: 1) Deceptive packaging - presented as agent skill but contains no implementation code, 2) Hidden data exfiltration - user source code is sent to third-party API without prominent disclosure, 3) Missing security metadata - no manifest fields indicating external dependencies or network access. While not malicious in intent, this skill could lead users to inadvertently expose sensitive code to external services. The skill should either include actual implementation code or clearly disclose its nature as an external tool wrapper with explicit warnings about data transmission.
Scanner's overall assessment (LLM analyzer)

Findings2

Critical0

High0

Medium2

Low0

Analyzersstatic behavioral llm trigger meta

Second scanner: NVIDIA SkillSpector

VerdictSafe 12

Score12 / 100

Issues5

CategoriesMCP Rug Pull

Version2.12.0 (static)

Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.

What this skill does

Fix Core Web Vitals issues with AI guidance. Use when your Lighthouse scores are bad.

Package: 1 file, 2.1 KB
SKILL.md: 307 words, 12+ sections
Code: None. Instructions only.

SKILL.md covers: Core Web Vitals Fixer · Quick Start · What It Does · Usage Examples · Scan your source directory · Scan app directory · Focus on specific metric · Best Practices · When to Use This · Part of the LXGIC Dev Toolkit · Requirements · How It Works

Findings

MEDIUM Misleading Skill Description - External Tool Masquerading as Agent Skill Social Engineeringllm

The skill presents itself as an agent skill for fixing Core Web Vitals issues, but the instructions describe an external NPX package ('npx ai-core-vitals') that requires OPENAI_API_KEY and makes external API calls. This is deceptive because: 1) Agent skills should contain executable code within the package, not just instructions to run external tools, 2) The skill provides no actual implementation files, 3) Users may believe they're using a local agent capability when they're actually executing an external package that sends code to OpenAI's API, 4) No disclosure about data transmission to third-party services.

Instructions state 'npx ai-core-vitals ./src/' and 'Needs OPENAI_API_KEY environment variable' and 'sends them to GPT-4o-mini' - indicating this is an external tool, not a self-contained agent skill

Remediation: Either: 1) Include actual implementation code in the skill package to make it a true agent skill, OR 2) Clearly disclose in the description that this is a wrapper for an external NPX package that sends code to OpenAI's API, OR 3) Reclassify this as documentation rather than an executable skill

MEDIUM Undisclosed External Data Transmission to Third-Party API Data Exfiltrationllm

The skill instructions reveal that user code is sent to OpenAI's GPT-4o-mini API ('sends them to GPT-4o-mini') without prominent disclosure in the description or manifest. This constitutes potential data exposure because: 1) User source code is transmitted to external servers, 2) The YAML description makes no mention of external API calls or data transmission, 3) Users may inadvertently expose proprietary code, credentials, or sensitive business logic, 4) No consent mechanism or privacy warning is provided before code transmission.

'Scans your code files for common performance antipatterns, then sends them to GPT-4o-mini' - code is transmitted externally without clear upfront disclosure

Remediation: 1) Add prominent disclosure in YAML description about external API usage and data transmission, 2) Warn users that source code will be sent to OpenAI, 3) Implement opt-in consent mechanism, 4) Provide guidance on excluding sensitive files, 5) Document data retention and privacy implications

What is in the package

File	Kind	Size
`SKILL.md`	doc	2.1 KB

External hosts referenced

github.com x.com lxgicstudios.com lxgicstudios.substack.com

Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.

Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.

What we checked

Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.

What we did not check

Runtime behaviour. The skill was not executed; findings come from reading its files.
Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
Content the analyzers cannot read: images, binaries, and some non-English text.
Known scanner bypass classes. See the methodology and limitations page.

Sources

Repository: https://clawhub.ai/skills/ai-core-vitals
Raw scan result: core-web-vitals-optimizer-scan.json
Badge for your README: ![skillscan.io](https://skillscan.io/badge/clawhub-ai-core-vitals-fixer.svg)
Request a re-scan