The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation in SKILL.md includes standard instructions for installing dependencies using npm and pip. These represent routine development setup procedures.
[COMMAND_EXECUTION]: The toolkit provides Python scripts (scripts/architecture_diagram_generator.py, scripts/project_architect.py, and scripts/dependency_analyzer.py) intended to automate architectural tasks. These scripts perform target path validation and generate status reports without executing arbitrary or malicious shell commands.
[CREDENTIALS_UNSAFE]: The skill documentation follows security best practices by instructing users to manage environment variables via a .env file template (.env.example) rather than hardcoding sensitive credentials.
[INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection because it is designed to ingest and process local project directories and has access to exploitable capabilities like package managers and container tools.
Ingestion points: Target path arguments passed to scripts in the scripts/ directory as described in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded content are present in the provided script skeletons.
Capability inventory: The skill instructions reference the use of npm, pip, docker, docker-compose, and kubectl for project management and deployment.
Sanitization: The scripts do not currently implement sanitization or filtering of the target path or its contents.

senior-architect