The Agent Skills Directory

[SAFE]: The skill is designed to operate locally on source code files. No patterns of data exfiltration or unauthorized network communication were found.\n- [SAFE]: Command execution is restricted to standard version control tools (git) and the provided local Python scripts used for structural code analysis.\n- [SAFE]: The included Python scripts (code_analyzer.py, simplification_rules.py, and change_reporter.py) use standard library modules and regex-based logic to evaluate code complexity without the use of dynamic execution or untrusted imports.\n- [SAFE]: No obfuscation, hardcoded credentials, or unauthorized persistence mechanisms were identified in the instructions or scripts.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection attack surface because it processes untrusted source code as data.\n
Ingestion points: Project source files (e.g., .ts, .py) identified via git diff in SKILL.md.\n
Boundary markers: The instructions do not specify explicit delimiters or 'ignore' instructions for the code content being refactored.\n
Capability inventory: The agent has access to Edit, Write, and Bash tools to modify the local file system.\n
Sanitization: No specific content filtering or sanitization is performed on the ingested source code.\n
Note: This surface is assessed as SAFE as it is inherent to the functionality of a refactoring skill and is mitigated by the use of deterministic local analysis scripts.

nav-simplify