The Agent Skills Directory

[SAFE]: The skill does not contain any malicious code or hidden payloads. All security-sensitive terms (e.g., prompt injection phrases or credential keywords) are exclusively used within regex patterns for scanning external directories as part of its auditing function.\n- [COMMAND_EXECUTION]: The skill provides PowerShell scripts for generating and verifying SHA-256 hashes of files. These scripts use standard Windows cmdlets and are appropriate for the tool's goal of ensuring supply chain integrity.\n- [PROMPT_INJECTION]: While instructions like 'ignore all previous instructions' are present in the text, they are provided as examples of what to search for during an audit, rather than being part of a jailbreak or override attempt against the agent using the skill.\n- [CREDENTIALS_UNSAFE]: The skill mentions common secret-related keywords (TOKEN, SECRET, KEY) only to define detection rules for auditing environment variable usage in other packages.

agent-supply-chain