The Agent Skills Directory

[CREDENTIALS_UNSAFE]: Analyzed code snippets for hardcoded secrets. The skill properly uses placeholders such as 'sk_...' and 'pk_...' for Stripe API keys and 'encryption_key' variables, which is a standard and safe practice for providing implementation templates.
[DATA_EXFILTRATION]: Evaluated data handling logic. The skill promotes secure practices like PAN masking and explicitly prohibits storing sensitive card data (CVV, PIN). No network exfiltration or unauthorized file access patterns were found.
[DYNAMIC_EXECUTION]: Inspected code for unsafe evaluation or deserialization. The implementation of tokenization and encryption utilizes well-known, secure libraries like cryptography and secrets. No risky dynamic execution patterns such as eval() or unsafe pickle loads were identified.
[COMMAND_EXECUTION]: Scanned for subprocess calls and shell commands. No instances of arbitrary command execution or risky system calls were found in the provided instructions or scripts.
[EXTERNAL_DOWNLOADS]: Checked for remote dependencies or script fetches. The skill mentions standard Python libraries and does not include commands to download or execute external, unverified code.

pci-compliance