update-threat-db

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The provided threat-db.yaml documents multiple confirmed malicious skills, supply-chain backdoors, and IOCs (reverse shells, credential-exfil endpoints, RAT behaviors and clipboard/secret theft), indicating deliberate malicious intent and backdoor activity.

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

• Third-party content exposure detected (medium risk: 0.30). The “Update Threat Database” skill ingests outsider-authored free text by performing targeted Perplexity (or WebSearch) lookups over security/news pages (Phase 2), then uses the returned findings as LLM input for analysis and deduplication.

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

• Secret detected (high risk: 1.00). I scanned the SKILL.md and threat-db.yaml content for literal, high-entropy credentials that would grant access to a service.
• Flagged: webhook.site/358866c4-81c6-4c30-9c8c-358db4d04412 — this is a literal webhook.site URL containing a high-entropy token/UUID that grants access to the webhook inbox (can be used to view/receive captured requests). It is a real, usable endpoint-like secret and therefore should be treated as sensitive.
• Ignored items:
• "openclaw" (password: 'openclaw') — low-entropy, known setup/archive password used in delivery examples; treated as a non-sensitive example per rules.
• Truncated hashes like "1e6d4b05...e2298" — redacted/truncated values, ignore.
• IP addresses, domains, and CVE IDs — IOCs and indicators, not credentials.
• No API keys (sk-, ghp_, AKIA*, etc.), private keys, or other high-entropy API tokens were present elsewhere in the content.

Therefore I mark a real secret present (the webhook.site URL) and list its locations below.