The Agent Skills Directory

[SAFE]: The skill executes read-only git commands (git status, git diff, git log) to assess the current state of the repository for summary purposes. These are routine operations for a development-focused agent skill.
[SAFE]: The skill provides options to save summaries to local files or export them using platform-integrated features (e.g., Claude Code's export-summary). These actions are user-controlled and do not involve unauthorized data exfiltration to external domains.
[SAFE]: Data ingestion occurs via session history and git outputs; however, the skill focuses on summarizing this content without executing it, minimizing risks associated with indirect prompt injection. No sanitization logic is explicitly defined, but the output is intended for user review or documentation.
[SAFE]: No evidence of credential harvesting, unauthorized network activity, or malicious obfuscation was found in the provided files.

session-wrap