neuropixels-analysis

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill implements standard neuroscience workflows using SpikeInterface and provides robust documentation and runnable scripts. No malicious intent, prompt injection, or obfuscation techniques were identified.
  • [CREDENTIALS_UNSAFE]: The skill demonstrates safe credential management practices. In both SKILL.md and references/AI_CURATION.md, it explicitly warns against hardcoding API keys and correctly demonstrates reading sensitive tokens (e.g., ANTHROPIC_API_KEY, OPENAI_API_KEY) from environment variables (os.environ).
  • [EXTERNAL_DOWNLOADS]: The skill lists standard, verifiable scientific Python packages (spikeinterface, probeinterface, kilosort, bombcell, etc.) from official package registries or public GitHub repositories of well-known community tools. It also references loading pretrained UnitRefine models from the official SpikeInterface Hugging Face repository, while properly cautioning the user to only trust verified model sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:45 AM
Security Audit — agent-trust-hub — neuropixels-analysis