The Agent Skills Directory

[SAFE]: The skill provides structured workflows for enumerating and auditing cloud resources using official SDKs and command-line interfaces.
[SAFE]: The deterministic detection of an 'eval/exec' pattern is a false positive; the skill uses grep to search for these strings in target source code as part of a security vulnerability assessment and does not execute them within its own process.
[SAFE]: The skill handles sensitive metadata such as IAM policies and environment variables locally for reporting; no network exfiltration routes to untrusted external domains were found.
[SAFE]: Dependency usage is limited to established libraries like boto3 and industry-standard security scanners like Bandit, Prowler, and Checkov.

performing-serverless-function-security-review