splits
Pass
Audited by Gen Agent Trust Hub on Jul 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements secure credential management by instructing the agent to use environment variables (
SPLITS_API_KEY) or standard input for authentication, ensuring that API keys are not stored in the shell history. - [SAFE]: Local sensitive data, including agent signer keys, are stored in the user's home directory with restricted filesystem permissions (
0600) to prevent unauthorized local access. - [SAFE]: The skill mandates a human-in-the-loop security model by defaulting to multisig thresholds of 2 or higher for agent-signed accounts, requiring human approval for transaction execution.
- [SAFE]: Explicit defensive instructions are provided to mitigate indirect prompt injection risks by requiring the agent to treat all external contract metadata, ABIs, and calldata as untrusted until verified against canonical sources.
- [EXTERNAL_DOWNLOADS]: The skill installs the official Splits CLI package (
@splits/splits-cli@0.2.9) from the NPM registry using a pinned version, which is a standard procedure for this integration and targets a well-known service.
Audit Metadata