segyio

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted SEG-Y files and extracts text headers which are printed to the agent's context. This creates a surface for indirect prompt injection where malicious instructions embedded in the seismic metadata could influence the agent's behavior. \n
  • Ingestion points: The scripts/inspect_segy.py script reads the text header from the input file at line 43 (info['text_header'] = f.text[0]). \n
  • Boundary markers: Absent. The text header content is printed directly without delimiters or warnings to the agent. \n
  • Capability inventory: The skill has the capability to write files to the local filesystem via scripts/extract_subset.py using segyio.create. \n
  • Sanitization: Absent. There is no sanitization or validation of the text header content before it is processed or displayed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 11:56 AM
Security Audit — agent-trust-hub — segyio