essential-mcp-servers-reference

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt contains numerous explicit-looking API key placeholders (eg. ghp_, AKIA..., sk_test_/sk_live_) and shows commands that print or embed tokens (echo $GITHUB_TOKEN, curl -H "Authorization: Bearer $GITHUB_TOKEN"), which would enable or encourage an agent to handle and output secret values verbatim — a high exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill contains many npx install commands that fetch and execute remote npm packages at runtime (e.g., "npx -y @modelcontextprotocol/server-github"), which downloads and runs external code that the skill relies on to add MCP servers.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The document explicitly lists and provides installation/configuration examples for payment, banking, trading, and blockchain servers (e.g., Stripe, PayPal, Plaid, Alpaca, CCXT, Base, Solana, thirdweb). It includes guidance on supplying secret keys, sandbox vs live keys, and notes that these servers can handle real money or execute irreversible on-chain transactions. These are specific tools/APIs whose primary purpose is moving or managing funds, not generic interfaces. Therefore the skill grants Direct Financial Execution Authority.