cryptotokenkit
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The documentation provides an example command using
sudoto register the token extension with the system's security agent. This is a standard and documented administrative procedure for installing CryptoTokenKit extensions. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from external hardware (smart cards) using APDU commands and TLV parsing, which represents a boundary where untrusted data enters the application context.
- Ingestion points: Smart card responses captured via
TKSmartCard.sendinSKILL.mdandreadPIVObjectinreferences/cryptotokenkit-patterns.md. - Boundary markers: Not explicitly implemented; logic assumes hardware integrity.
- Capability inventory: Keychain access and cryptographic operations (signing, decryption, and key exchange).
- Sanitization: Structured parsing using
TKBERTLVRecordis employed, though further semantic validation of the parsed data content is recommended as a best practice.
Audit Metadata