contentsquare-web-tag-install

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly requires the developer's Contentsquare tag ID to be embedded verbatim into generated code and CLI commands (clientId and --tag-id), forcing the LLM to handle/output a secret-like identifier.

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

• Third-party content exposure detected (medium risk: 0.30). Runtime verification is done by the CLI npx --yes @contentsquare/wizard@2 verify which launches a real browser to the developer-provided --url and reports CSP issues and navigation/pageview evidence, meaning it ingests free text/URLs from the visited web app at runtime (outsider-authored content path: the site loaded at --url).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes npx --yes @contentsquare/wizard@2 (which downloads and executes remote code at runtime) and requires loading the remote tag script https://t.contentsquare.net/uxa/<TAG_ID>.js in the browser during verification, so remote code is fetched and executed as part of the runtime flow.