fittings-doc-html-to-markdown

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs exporting a private key into the shell (FITTINGS_PRIVATE_KEY=0x...) and notes the key “sits in plaintext wherever the agent runs,” which encourages handling secrets directly and could lead an agent to echo or embed sensitive key material verbatim in commands or outputs.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). Most URLs are API/docs on a single domain and look legitimate, but https://fittings.sh/skill/call.mjs is a downloadable JavaScript the skill instructs you to curl and run (including passing a wallet private key), which is a high-risk pattern—downloading and executing third-party scripts that handle secrets.

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow reads the outsider-supplied html field from the POST request body to https://fittings.sh/v1/doc/html-to-markdown and processes that HTML text for conversion.

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

• Direct money access detected (high risk: 1.00). The skill describes a paid API that uses an on-chain/crypto payment flow: it requires a wallet holding USDC on Base, shows setting FITTINGS_PRIVATE_KEY, and instructs running a client (call.mjs) that signs an authorization so a facilitator settles the payment. It even references @x402/evm and off-chain signing. Those are explicit crypto payment instructions (wallet/private-key signing to authorize settlement), which constitute direct financial execution capability.