The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches resources from external domains during installation and setup.\n
Evidence: curl -s https://ethskills.com/wallets/SKILL.md and claude plugin install https://github.com/austintgriffith/ethskills in catalog.json.\n- [COMMAND_EXECUTION]: The skill specifies shell commands for installation and setup.\n
Evidence: curl and platform-specific install commands in catalog.json.\n- [PROMPT_INJECTION]: The skill directs the agent to fetch and process instructions from a remote URL, establishing an indirect prompt injection surface (Category 8).\n
Ingestion points: The agent is instructed to read content from https://ethskills.com/wallets/SKILL.md.\n
Boundary markers: There are no delimiters or warnings to prevent the agent from treating remote content as high-priority instructions.\n
Capability inventory: The skill handles Ethereum transaction logic, which increases the impact if the fetched content is malicious.\n
Sanitization: No validation or sanitization of the remote content is performed before the agent processes it.

Wallets