cross-team-comm

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). High-risk: the document provides step-by-step instructions to remotely access other teams' OpenClaw instances, collect tokens/credentials from their files, change bindings to enable remote access, and use gateway tokens to control agents—patterns consistent with credential theft and unauthorized remote takeover.

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

• Secret detected (high risk: 1.00). I inspected the skill content for high-entropy literal credentials. I ignored obvious placeholders and examples such as "", "<app_id>", "<app_secret>", "$(pass show ...)" and other template values because they are documentation placeholders or references to secret storage.

However, line 34 contains "REDACTED_FEISHU_GROUP_TOKEN_2", which is a redaction marker indicating a real credential was present and stripped. Per the provided rules, literal redaction markers of this form should be treated as evidence of a secret and therefore flagged.