The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local bash script (scripts/validate.sh) to perform architectural analysis. This script utilizes standard Unix utilities like find and grep to scan the codebase for import violations. The execution is scoped to the local filesystem and intended for code quality validation.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads content from external Python files and returns the results to the agent's context. A malicious file could include instructions disguised as comments or code that attempt to override the agent's behavior when reviewing the validation report.
Ingestion points: The scripts/validate.sh script reads any .py file located within the src/project_watch_mcp/ directory structure.
Boundary markers: No specific boundary markers or 'ignore' instructions are used to wrap the output returned by the bash script.
Capability inventory: The skill is authorized to use Read, Grep, and Bash tools, allowing it to inspect files and execute scripts.
Sanitization: The validation script does not sanitize or escape the content of the lines it extracts using grep before presenting them to the agent.

architecture-validate-layer-boundaries