ui-test
Verdict safe, but 2 findings (1 medium, 1 low), mainly unauthorized tool use. Most recent scan 2026-02-07.
The ui-test skill has MEDIUM severity issues related to incomplete implementation and missing tool declarations. The skill's core functionality depends on external scripts (ui-test.js) that are not included in the package, creating a broken dependency. Additionally, the skill uses browser automation tools extensively but does not declare these dependencies in allowed-tools, reducing transparency. There are no critical security threats like data exfiltration, prompt injection, or credential theft. The skill's purpose (UI testing via browser automation) is legitimate and matches its description. The primary concerns are completeness and transparency rather than malicious intent.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
The two scanners disagree: Cisco found nothing and SkillSpector says do not install. One of them has a blind spot here; read the skill before trusting either. How the two scanners compare across the whole corpus: Two scanners, one corpus.
What this skill does
Plain English E2E UI testing. Describe tests in natural language, agent executes via browser tool, then exports real Playwright test scripts for CI/CD. Use when asked to create UI tests, test a website, or generate Playwright scripts.
- Package
- 2 files, 19 KB
- SKILL.md
- 701 words, 8 sections
- Code
- JavaScript (1)
SKILL.md covers: UI Test — Plain English E2E Testing (🧪) · Workflow · Agent Execution Flow · Step Interpretation Guide · Commands · Export Format · Screenshots & Video · Storage
Findings
MEDIUM Missing Script Implementation - ui-test.js Not Provided Unauthorized Tool Usellm
The instructions reference 'node scripts/ui-test.js' multiple times for core functionality (get, save-run commands) but no JavaScript files are included in the skill package. This creates a broken dependency where the skill cannot function as described, potentially leading to unexpected behavior or errors when the agent attempts to execute the documented workflow.
SKILL.md
Instructions reference:
- 'node scripts/ui-test.js get "<name>"'
- 'node scripts/ui-test.js save-run "<name>" passed=true/false'
But: 'No script files found.'LOW Undeclared Tool Usage - Browser Tool Not Listed in allowed-tools Unauthorized Tool Usellm
The skill extensively uses the 'browser' tool (browser action=start, browser action=snapshot, browser action=act, browser action=screenshot) but does not declare any allowed-tools in the YAML manifest. While allowed-tools is optional, the skill's heavy reliance on browser automation without explicit declaration creates ambiguity about intended tool permissions and could mask unauthorized tool usage.
SKILL.md
allowed-tools: Not specified
But instructions require: 'browser action=start profile=clawd', 'browser action=snapshot', 'browser action=act', 'browser action=screenshot'What is in the package
| File | Kind | Size |
|---|---|---|
SKILL.md | doc | 5.2 KB |
scripts/ui-test.js | script | 14 KB |
External hosts referenced
myapp.com files.catbox.moe pleblab.dev
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.
- Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
- 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/ui-test
- Raw scan result: ui-test-scan.json
- Badge for your README:
- Request a re-scan