The Agent Skills Directory

[COMMAND_EXECUTION]: The helper scripts loc_counter.py and git_session_analyzer.py execute git commands via subprocess.run. These calls use list-based arguments and avoid shell interpretation, making them safe from command injection.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) as it processes external repository data.\n
Ingestion points: File paths and commit subjects are read into the agent's context during analysis by loc_counter.py and git_session_analyzer.py.\n
Boundary markers: The skill does not implement explicit delimiters or warnings between the analyzed data and the final report output.\n
Capability inventory: Local file reading and git command execution are the primary capabilities across the helper toolchain.\n
Sanitization: The scripts use basic error replacement for decoding issues during file operations, though they do not actively filter for instruction-like strings in metadata.\n- [SAFE]: No hardcoded credentials, remote downloads, or unauthorized network operations were detected in the skill instructions or scripts. All operations trace back to legitimate repository analysis functionality.

cost-estimate