The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard version control and development tools, including git log, git diff, and git branch, to align its persistent memory files with the current project state. These operations are transparently documented as part of the core functionality.\n- [EXTERNAL_DOWNLOADS]: The included utility script scripts/token-audit.py references the tiktoken Python library for tokenization. This is a well-known, reputable library from OpenAI, and the script includes a fallback mechanism if the package is not installed.\n- [DATA_EXFILTRATION]: While the skill reads project metadata and developer preferences to maintain context, all data is strictly stored locally in the project root or the user's home directory (~/.claude/). The skill contains explicit rules against storing credentials or sensitive information.\n- [PROMPT_INJECTION]: The skill body consists of instructional guidelines for memory management. It does not contain any patterns indicative of safety filter bypass, system prompt extraction, or adversarial role-play.

memory-bank