NYC
skills/letta-ai/skills/sam-cell-seg/Gen Agent Trust Hub

sam-cell-seg

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No instructions targeting agent behavior override, role-play injection, or safety filter bypass were detected. The instructions are focused on technical implementation and interface matching.
  • [Data Exposure & Exfiltration] (SAFE): The skill involves processing local images and CSV files for coordinate extraction. No network operations to unknown domains or access to sensitive system paths (e.g., SSH, credentials) were found.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): The code relies on standard machine learning libraries (PyTorch, OpenCV, NumPy). No suspicious external scripts or unverified package installations are suggested.
  • [Dynamic Execution] (SAFE): The skill contains a testing snippet using subprocess.run to check the script's help menu. This is a standard development practice and is implemented safely using the local Python executable with static flags.
  • [Obfuscation] (SAFE): No hidden, encoded, or deceptive content (Base64, zero-width characters, or homoglyphs) was identified in the code or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:04 PM