ctx-stats

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt mandates copying the ENTIRE tool output verbatim into the reply, so if the MCP output contains any API keys, tokens, cookies, or other secrets (or raw user content), the agent is forced to include them in its output, enabling secret exfiltration.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill contains a prompt-injection style requirement forcing verbatim, unredacted copy-paste of an MCP tool's full output (no summarization or redaction), which creates a clear data-exfiltration risk of leaking tokens, credentials, or other sensitive internal information.