The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill includes a setup step that downloads and executes an installation script from 'https://tempo.xyz/install'. While this involves remote code execution, it is the standard installation method for the Tempo CLI, which is the primary tool the skill is designed to manage. This risk is further mitigated by the skill's instruction to the agent to explain steps to the user.
[COMMAND_EXECUTION]: The skill uses the 'Bash' tool to execute various CLI utilities including 'tempo', 'cast', and 'jq'. These are used for wallet management, blockchain interaction, and data processing. The skill mandates explicit user confirmation via 'AskUserQuestion' before any transaction-related command is run.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from HTTP 402 payment challenges (headers and JSON bodies) to determine payment requirements.
Ingestion points: Payment details are extracted from 'WWW-Authenticate' headers and 402 response bodies.
Boundary markers: The skill contains a 'REJECT' rule for any value containing shell metacharacters (e.g., ;, |, &, $), creating a clear boundary between data and execution.
Capability inventory: The agent has access to 'Bash', 'curl', 'cast', and 'tempo' for network and blockchain operations.
Sanitization: The skill defines 'Input Validation Rules' using regex for Ethereum addresses, chain IDs, and token amounts to ensure only safe, well-formatted data is processed.
[EXTERNAL_DOWNLOADS]: The skill references several external resources including 'tempo.xyz', 'trade-api.gateway.uniswap.org', and 'mpp.dev'. These are the official endpoints for the services the skill integrates and are used for tool installation and API communication.

pay-with-any-token