onchain-alpha
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-provided CSV files. While this creates an ingestion point for potential indirect prompt injection, the accompanying diagnostics script implements robust sanitization by strictly coercing data into numeric and datetime formats. * Ingestion points: scripts/onchain_alpha_diagnostics.py reads the input_csv argument. * Capability inventory: The script is limited to data summarization and local file writing; it contains no network, shell execution, or privilege escalation capabilities. * Boundary markers: None. * Sanitization: High; uses pandas numeric and datetime coercion to neutralize non-numeric content.
- [EXTERNAL_DOWNLOADS]: The skill depends on the pandas library for its diagnostics logic, which is a standard and trusted package for data analysis tasks.
Audit Metadata