skills/duneanalytics/skills/sim/Gen Agent Trust Hub

sim

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill leverages the Bash tool to run dune CLI commands and curl for API interactions. This is the intended functionality for performing on-chain lookups.
  • [EXTERNAL_DOWNLOADS]: The documentation mentions the automatic installation of the Dune CLI. This is a known vendor resource from duneanalytics and is required for the skill's operation.
  • [CREDENTIALS_UNSAFE]: The skill includes explicit warnings against passing sensitive API keys via command-line flags, instead promoting the use of environment variables and configuration files to prevent credential exposure.
  • [DATA_EXFILTRATION]: Instructions are provided for the agent to automatically redact sensitive strings (such as those with a sim_ prefix) from its output, providing a layer of protection against accidental data leakage.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection. 1. Ingestion points: Remote on-chain data retrieved via 'dune sim' commands from the Dune Sim API. 2. Boundary markers: The agent is instructed to use the '-o json' flag for structured, machine-readable responses. 3. Capability inventory: Access to 'Bash' for CLI command execution and 'curl' for network requests. 4. Sanitization: The agent is required to redact API keys and sensitive tokens from the output before presenting it to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 02:32 PM