The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands via backticks to resolve directory paths ($UA_DIR) and verify Git commit status.
[COMMAND_EXECUTION]: Executes several git commands (git rev-parse, git diff, git ls-files) to compare the knowledge graph's metadata with the current repository state to identify stale data.
[COMMAND_EXECUTION]: Utilizes grep to perform targeted searches within a local knowledge-graph.json file rather than loading the entire file into the model's context, which is a recommended performance practice.
[SAFE]: All file access and command execution are restricted to the local project directory and the user-provided arguments, with no evidence of remote data exfiltration or unauthorized access to sensitive files like SSH keys or credentials.

understand-explain