batch-token-price-lookup
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override agent behavior, bypass safety guidelines, or extract system prompts were detected. The triggers and instructions are strictly aligned with the stated purpose of price lookups.
- [DATA_EXFILTRATION]: The skill does not access sensitive local files (e.g., SSH keys, credentials) or environment variables. No hardcoded secrets were found. All network operations are performed through defined DexPaprika MCP tools.
- [EXTERNAL_DOWNLOADS]: The skill does not download external scripts, install packages, or fetch remote code for execution.
- [REMOTE_CODE_EXECUTION]: No use of dangerous functions such as eval(), exec(), or subprocess calls with untrusted input was identified.
- [COMMAND_EXECUTION]: No unauthorized shell commands or privilege escalation attempts were found.
- [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it processes user-provided token addresses and network names. Ingestion points: User input and URLs containing token addresses. Boundary markers: The skill uses structured formatting for its output to separate data from instructions. Capability inventory: Token and network metadata lookups via getTokenMultiPrices and getTokenDetails tools. Sanitization: The skill explicitly includes a validation and normalization stage to verify addresses and network synonyms before tool invocation.
Audit Metadata