nansen-pm-insider-scan
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill leverages the
nansenresearch CLI to perform data retrieval and analysis. It executes commands such asmarket-screener,pnl-by-market, andhistorical-balancesto query blockchain data from the Polygon network. - [PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection where user-provided data is used in a shell command query.
- Ingestion points: The
<market name>placeholder in thenansen research prediction-market market-screenercommand (SKILL.md). - Boundary markers: The user-supplied query is enclosed in double quotes within the shell command template.
- Capability inventory: The skill initiates shell commands to interact with external APIs via a specialized CLI tool.
- Sanitization: The skill provides no explicit sanitization for the market name input, relying on the agent's internal safety measures and shell-level quoting.
Audit Metadata