sim-studio

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt embeds a plaintext database URL with a real-looking password and shows/instructs placing API keys into headers/config (e.g., X-API-Key:{YOUR_API_KEY}), which requires including secrets verbatim and is therefore insecure.

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md's "Workflow Patterns for Automation" (Pattern 1: Research Pipeline — "Web Search > Content Extract > LLM Summarize > Output") explicitly describes deployed MCP workflows that fetch and ingest public web content which Claude Code will read and act on, creating a clear path for untrusted third-party content to influence agent behavior.

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

• Secret detected (high risk: 1.00). The PostgreSQL connection string in the "Local Instance" table contains a literal username and password: "postgres:simstudio_local_2026" (postgresql://postgres:simstudio_local_2026@localhost:5433/simstudio). This is a directly present, usable credential and therefore flagged.

Other items mentioning API keys or admin keys are either references/placeholders (e.g., "Admin API Key is in ~/Projects/sim/.env", "{YOUR_API_KEY}", "$ADMIN_API_KEY", or header examples) or local host URLs and ports — those are documentation placeholders or environment references and do not contain literal secrets, so they are ignored.