wema-bmc

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt embeds an explicit service-account username and plaintext password (and shows curl/python examples that include them and the resulting JWT), which forces the agent to handle and potentially emit secret values verbatim in commands and code.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The document contains live service-account credentials, full API endpoints, and ready-to-run scripts that enable direct access to sensitive AR System data (people, incidents, changes) and therefore provides a clear capability for unauthorized data access/exfiltration and abuse.

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

• Secret detected (high risk: 1.00). I scanned the skill content for high-entropy, literal credentials that would grant access.

Findings:

• The document includes a service account username and password used to obtain JWT tokens:
• Username: zedai@wemabank.com (identifier, not a secret by itself)
• Password: "Owzre !Er,s3-37" — present in the Service Account table, in curl examples, and in the Python helper (PASS = "Owzre !Er,s3-37"). This is a literal credential (complex/mixed characters) that can be used to authenticate to the live REST API, so it meets the definition of a secret.

No other high-entropy secrets (API keys, private keys, or redaction markers) were present. URLs, emails, form names, example incident IDs, and simple words were treated as non-secrets per the ignore rules.