The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute rg (ripgrep) commands to perform static analysis of the local project directory. These commands are used to detect hardcoded secrets, dangerous functions (like shell=True or eval()), and configuration errors (like DEBUG=True). These operations are performed locally and are consistent with the skill's stated purpose.
[PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from the user's project files, which introduces a surface for indirect prompt injection.
Ingestion points: The skill scans the entire project structure, including source code files (.py, .js, .go, .java), configuration files (.env, .yaml, .json), and infrastructure manifests (Dockerfile, Kubernetes manifests).
Boundary markers: The instructions do not specify the use of boundary markers or clear delimiters to separate untrusted file content from the agent's internal reasoning or audit logic.
Capability inventory: The skill has the capability to read files across the project directory and execute shell commands (rg) to search through them.
Sanitization: There is no evidence of sanitization or escaping mechanisms applied to the content read from files before it is analyzed or included in the final audit report.

production-check