The Agent Skills Directory

[COMMAND_EXECUTION]: The scaffold-endpoint and scaffold-ai-endpoint commands perform filesystem write operations to generate and save new project source code to user-specified directories.
[EXTERNAL_DOWNLOADS]: The execute-endpoint and probe-endpoint subcommands allow for making network requests to any user-provided HTTPS URL, which may lead to interactions with untrusted external services.
[CREDENTIALS_UNSAFE]: The send-inbox-message subcommand retrieves the wallet's private key to sign and serialize sBTC transactions for message delivery.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through external data ingestion.
Ingestion points: JSON responses from arbitrary external API endpoints accessed via the execute-endpoint command.
Boundary markers: None identified; API responses are printed and potentially processed without delimiters.
Capability inventory: Filesystem writes (scaffolding), network requests (fetch), and transaction signing (wallet interaction).
Sanitization: None; data from external sources is parsed and returned without validation or escaping.

x402