The Agent Skills Directory

[SAFE]: The skill consists of analytical scripts and technical documentation for system design. The Python scripts (architecture_diagram_generator.py, dependency_analyzer.py, and project_architect.py) perform static analysis on local project structures using standard libraries for file system traversal and regex-based parsing. No findings related to credentials exposure, remote code execution, persistence, or privilege escalation were detected.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external project data to produce architectural assessments and diagrams, which creates a potential surface for indirect prompt injection.
Ingestion points: Source code, package manifests (package.json, requirements.txt, go.mod, Cargo.toml), and configuration files are read from the user-provided project directory by the tools in the scripts/ directory.
Boundary markers: The tools do not utilize explicit delimiters or "ignore embedded instructions" warnings when outputting the analysis results to the agent.
Capability inventory: The tools are limited to reading local files and writing report outputs. They lack network capabilities, shell execution, or privileged operations.
Sanitization: The scripts perform static analysis only (regex matching and structural metrics) and do not execute or interpret the contents of the files they read.

senior-architect