a11y-self-check

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill facilitates accessibility audits by writing generated code to temporary files and executing a local shell script (lint-html.sh) for analysis. This is a standard development workflow and does not involve malicious operations.
  • [COMMAND_EXECUTION]: Instructs the agent to execute a local bash script located within the plugin directory (${CLAUDE_PLUGIN_ROOT}/skills/html-lint-runner/scripts/lint-html.sh) to process code written to temporary files in /tmp/.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for processing untrusted data.
  • Ingestion points: Code snippets are written to /tmp/claude-generated.html and /tmp/claude-generated.tsx in SKILL.md.
  • Boundary markers: Absent.
  • Capability inventory: Shell script execution (bash) and file system writes (cat).
  • Sanitization: No explicit sanitization is performed on the code content before it is passed to the linter script.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 04:46 AM
Security Audit — agent-trust-hub — a11y-self-check