core-nfc

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill contains multiple reference links to documentation on the domain sosumi.ai. This domain is not an official Apple resource or a recognized trusted vendor. Evidence includes documentation links in the References section of SKILL.md such as https://sosumi.ai/documentation/corenfc.
  • [PROMPT_INJECTION]: The skill provides implementation patterns for ingesting and processing data from external NFC tags. This creates an indirect prompt injection surface where malicious payloads on a tag could attempt to influence the logic of an agent or application processing the records.
  • Ingestion points: readerSession(_:didDetectNDEFs:) and NSUserActivity.ndefMessagePayload in SKILL.md, and tagReaderSession(_:didDetect:) in SKILL.md.
  • Boundary markers: Absent; tag payloads are processed directly without delimiters or instruction-guarding.
  • Capability inventory: The skill includes patterns for reading, writing, and locking NFC tags, as well as updating application state.
  • Sanitization: Absent; code snippets process record data (e.g., record.payload) without escaping or validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 08:52 PM