vision-framework
Fail
Audited by Socket on Mar 8, 2026
1 alert found:
Obfuscated FileObfuscated FileSKILL.md
HIGHObfuscated FileHIGH
SKILL.md
The Vision Framework skill presents a coherent, properly scoped toolkit for on-device computer vision tasks within iOS/macOS ecosystems. Capabilities align with the stated purpose (OCR, face/barcode detection, document scanning, segmentation, tracking, and Core ML integration). There are no questionable download/execute patterns, no credential exposure, and data flows remain within on-device processing with clear sinks for results. The risk profile is low to moderate, primarily around data privacy considerations and correct use of on-device APIs; no evidence of unauthorized data exfiltration or supply-chain concerns.
Confidence: 98%
Audit Metadata