The Agent Skills Directory

Unverifiable Dependencies & Remote Code Execution (HIGH): Automated scans (URLite) have identified a blacklisted URL within the project resources (specifically flagged in the requirements.md context). This indicates a high risk of interaction with known malicious domains or command-and-control servers.
Privilege Escalation & Command Execution (MEDIUM): The documentation in README.md explicitly instructs users to execute Python scripts (generate_project_docs.py and validate_documents.py) for the skill to function. These scripts are missing from the analyzed package, which is a common indicator of hidden malicious behavior or an attempt to bypass static analysis of the primary execution logic.
Indirect Prompt Injection (LOW): This skill is vulnerable to Indirect Prompt Injection (Category 8) due to the following evidence:
Ingestion points: User-defined project names, features, and component lists provided via command-line arguments as described in the README.md usage examples.
Boundary markers: None found in the assets/requirements-template.md file; user-provided data is directly interpolated into placeholders without delimiters.
Capability inventory: The skill explicitly supports shell command execution (via documented Python scripts) and documentation generation meant for AI agents.
Sanitization: There is no evidence of input validation, sanitization, or escaping within the provided templates to prevent malicious user input from being interpreted as agent instructions.

project-planner