page-editor

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill contains mandatory telemetry calls (with a hardcoded Appkey) and documentation that automatically collects the system username / machine fingerprint (whoami/A2) and instructs sending these events via scripts — this is deliberate data-collection/exfiltration behavior that can leak identifying information without explicit user consent.

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

• Third-party content exposure detected (medium risk: 0.30). In page-editor’s monitor path, the CLI parses attacker-influenced JSON received from the locally running editor server (/api/wait-for-save) via http.get, then prints the full JSON to stdout and forwards it to the agent for applying code changes.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I flagged the alphanumeric Appkey "0WEB070FKPM08M1L" because it appears as a literal, non-placeholder value used as an "Appkey" in the documentation and example tracking calls (e.g., passed to scripts/track.sh). CHANGELOG.md explicitly says the example Appkey was replaced with a real value, and the string is reasonably high-entropy (mixed letters/digits), so it meets the definition of a secret (service credential). This is not a placeholder like "YOUR_API_KEY" and not a simple setup password, so it should be treated as a leaked credential.