matrixscan-count-ios
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains instructions for the AI agent to mask license keys in the chat interface. Although flagged as a potential concealment pattern by automated detectors, this is a legitimate security practice designed to protect sensitive credentials from accidental exposure and does not represent an attempt to bypass agent safety guidelines.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or other secrets were found in the skill files. The skill uses standard placeholders like
-- ENTER YOUR SCANDIT LICENSE KEY HERE --and instructs the agent to use proper tools for credential retrieval. - [EXTERNAL_DOWNLOADS]: The skill references official Scandit SDK repositories on GitHub and documentation on scandit.com. These are verified vendor resources necessary for the integration of the SDK and do not involve untrusted third-party sources.
- [COMMAND_EXECUTION]: The code samples provided are standard Swift and Objective-C implementation patterns for iOS development. They do not contain any arbitrary command execution, unauthorized subprocess calls, or other dangerous system-level operations.
Audit Metadata