accesslint-diff

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The required workflow runs npx ... @accesslint/cli@latest "<url>" ... --format json, so any outsider-supplied URL (from $ARGUMENTS) can cause the CLI to fetch and ingest arbitrary page content/evidence into the LLM-visible JSON output, creating an indirect prompt-injection path.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes npx to fetch and execute remote npm packages at runtime (e.g., "npx -y @accesslint/chrome@latest" and "npx -y @accesslint/cli@latest"), which pulls and runs code from the npm registry.