The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches reputable security scanners (Trivy, Gitleaks, Grype, Syft, Gosec, and Local PHP Security Checker) from official GitHub repositories. The skill implements security best practices by using pinned versions and verifying all binary downloads with SHA-256 checksums.
[COMMAND_EXECUTION]: Generates and executes Docker commands and shell scripts to orchestrate the auditing process. It also invokes project-native tools using user-specified prefixes. These operations are core to the skill's purpose and include isolation measures such as read-only filesystem mounts for the analyzed source code.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it analyzes untrusted source code and configuration files. Ingestion points: Project files are read during discovery and manual review phases. Boundary markers: None; the skill does not wrap analyzed code in delimiters or provide instructions to ignore embedded directives. Capability inventory: The agent can execute shell commands via Docker and write files to the project directory. Sanitization: No content filtering is applied to files before LLM analysis.

code-security-audit