srt-to-structured-data

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • SAFE (SAFE): No malicious patterns detected. The script performs legitimate file parsing and transformation consistent with its documentation.
  • Indirect Prompt Injection (SAFE): The skill possesses a surface for indirect prompt injection as it processes untrusted subtitle files. However, the script only performs structural text transformations and does not execute the parsed content, significantly mitigating the risk.
  • Ingestion points: scripts/parse_srt.py via the input file argument.
  • Boundary markers: Absent.
  • Capability inventory: Local file read (Path.read_text) and file write (Path.write_text) in scripts/parse_srt.py.
  • Sanitization: Content is parsed using rigid regular expressions to extract specific SRT components (index, timecodes, text).
  • Data Exposure & Exfiltration (SAFE): The script facilitates reading and writing files based on user-provided paths. While this is standard for a file-processing utility, it does not target sensitive system paths or perform any network operations to exfiltrate data.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:17 PM