Agentify
PassAudited by ClawScan on May 1, 2026.
Overview
Agentify is a coherent instruction-only skill for auditing and improving web markup, with no code, credentials, persistence, or hidden data flows shown.
This skill appears safe to install as an instruction-only web markup assistant. Before using it, be aware that it may read selected project files, fetch specified pages, and propose or apply template edits. Keep inputs scoped, review diffs carefully, and validate generated structured-data or accessibility changes before deployment.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent may read web pages and local source files selected by URL, path, glob, or IDE selection.
The skill directs the agent to use URL fetching and local file/glob access. This is expected for web-page analysis, but users should notice that invoking the skill may expose project files or fetched page content to the agent’s analysis context.
URL → fetch and retrieve HTML - File path → read the file - Glob pattern → find matching files, analyze each
Use specific paths or globs, avoid including unrelated sensitive files, and review any report or code changes before accepting them.
Accepted edits may change markup, metadata, accessibility attributes, and automation selectors in a website or application.
The skill can guide the agent to modify source templates. This is purpose-aligned and includes safety rules, but code edits to public web templates can affect production behavior or expose stable automation selectors if applied without review.
Transform web templates to be agent-friendly while preserving all existing functionality... Apply transformations in order: ... Add `data-testid` ... Insert JSON-LD structured data ... Add meta tags
Review diffs, run tests, and confirm that added selectors or structured data are intended before deploying changes.
If generated JSON-LD snippets are populated from untrusted content without proper escaping, they could introduce web security issues in the target application.
Framework guidance includes raw script/HTML insertion patterns for JSON-LD. This is common for structured data and aligned with the skill’s purpose, but generated snippets using dynamic values should be reviewed for proper escaping and safe handling.
dangerouslySetInnerHTML={{
__html: JSON.stringify({
"@context": "https://schema.org",
"@type": "Article",
"headline": title
})Ensure generated structured-data code safely serializes and escapes dynamic values, and have developers review framework-specific insertion patterns.