semantic-gap-analysis

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

• Third-party content exposure detected (low risk: 0.10). The workflow instructs the agent to fetch external web pages via search results (Google the target keyword and fetch top 3 competitor pages), which allows outsider-authored content to be ingested.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to fetch external URLs during execution (Fetch the URL, Google the target keyword. Fetch the top 3 results in full), creating a dynamic runtime external dependency with unknown provenance.