The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the installation of the @moonpay/cli package globally via npm. This is a vendor-owned resource associated with the skill author.
[CREDENTIALS_UNSAFE]: The skill interacts with sensitive local files such as ~/.config/moonpay/wallets.json and ~/.config/moonpay/credentials.json which store encrypted secrets. It also provides commands for the agent to handle highly sensitive data, including raw private keys and BIP39 mnemonics, during wallet import procedures.
[INDIRECT_PROMPT_INJECTION]: The "Autonomous login" section describes a workflow where the agent reads external email content using a third-party tool (gog) to automate OTP verification. This creates an attack surface where malicious instructions embedded in an email could be processed by the agent.
Ingestion points: External email content retrieved via gog gmail search as described in SKILL.md.
Boundary markers: Absent; the skill does not specify delimiters or instructions to ignore potential malicious content within the email body.
Capability inventory: The agent has the capability to manage wallets (create, import, delete) and perform authentication actions using the mp CLI.
Sanitization: No sanitization or validation of the incoming email data is mentioned before it is processed for the OTP code.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands via the MoonPay CLI to perform its primary functions.

moonpay-auth