The Agent Skills Directory

[SAFE]: No malicious patterns, credential exfiltration, or unauthorized behaviors were detected. The skill's functionality is well-aligned with its stated purpose of managing Solana wallet operations.- [COMMAND_EXECUTION]: The skill correctly uses the mirage and mirage ows CLI tools for legitimate blockchain activities, including wallet creation, balance inquiries, and transaction construction. These commands are documented for the intended workflow.- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect injection as it processes external Solana Program IDs and Anchor IDL files (ingestion points) to construct transactions. Its capabilities (capability inventory) include signing and sending these transactions. However, the skill provides clear boundary markers by instructing the agent to restate all transaction details and wait for explicit user confirmation (sanitization/safety check) before execution, mitigating the risk of following instructions embedded in malicious IDLs.

mirage-private-transfer