upload-and-share

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The workflow requires using the returned presigned uploadUrl (which contains a bearer-like token) verbatim in curl commands or outputs, so the agent would handle and potentially expose a secret-like value.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). These are public file‑hosting endpoints (stableupload.dev / f.stableupload.dev) that serve user-uploaded content — while the visible links are .png/.pdf, the service allows arbitrary unvetted files and can be used to distribute malware (and the workflow references running third‑party tooling), so it represents a moderate–high risk.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill mandates running the remote npm package via "npx agentcash@latest" at runtime (which fetches and executes code from the npm registry) to interact with the service (e.g., https://stableupload.dev/api/upload), so a required external package is executed and could control behavior.

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

• Direct money access detected (high risk: 1.00). The skill includes explicit payment and wallet operations: it instructs checking a USDC wallet balance with npx agentcash@latest wallet info, shows tiered prices, and provides a concrete purchase step using npx agentcash@latest fetch POST to https://stableupload.dev/api/upload to buy an upload slot (micropayment). These are specific crypto/payment actions (wallet balance, deposit link, spending USDC) rather than generic upload or HTTP instructions, so the skill grants direct financial execution capability.