solana-dev

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides instructions for installing the Solana/Agave toolchain using curl | sh from https://release.anza.xyz/stable/install.
  • [REMOTE_CODE_EXECUTION]: The skill suggests installing the Surfpool testing tool via curl | bash from https://run.surfpool.run/.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the official Solana Developer MCP server from https://mcp.solana.com/mcp to access live documentation and expertise.
  • [COMMAND_EXECUTION]: The skill utilizes several command-line tools for blockchain operations, including anchor, solana, avm (Anchor Version Manager), and surfpool. These are standard tools for the described development tasks.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes advanced security guardrails (W011) that explicitly warn the agent about untrusted on-chain data. It instructs the agent to ignore directives found in account metadata, token names, or program logs, which effectively mitigates indirect prompt injection attacks from malicious blockchain state.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 04:28 AM
Security Audit — agent-trust-hub — solana-dev