pp-public-param-golden

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The playbook replay instructions require substituting Playbook.slots_resolved entries (which the example shows can include a "token" / "") directly into commands, meaning the LLM would need to emit secret token values verbatim — an exfiltration risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). These links include a GitHub release and individual GitHub repos (mvanhorn / printing-press-golden) that the README directs users to download pre-built binaries/one-click bundles — downloading and running binaries from individual/unknown accounts is a common malware vector and therefore suspicious.

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow for the MCP tools includes a user-supplied search tool whose query argument is free text passed to handleSearch (i.e., outsider-authored text can be ingested into LLM runtime via the tool call path rather than only selecting pre-curated items).