The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill's logic is focused on legitimate business automation for the Zynkr organization.
[EXTERNAL_DOWNLOADS]: The skill documentation includes a command to fetch a skill-building utility from the author's own GitHub repository (github.com/peter-tu-zynkr). This is a legitimate vendor resource used for the installation and management of the skill.
[DATA_EXFILTRATION]: All data operations are confined to the user's authenticated Google Workspace environment for intended business purposes. The skill implements explicit logic in references/wb-entry-parsing.md to identify and protect sensitive morale scores, ensuring they are only reported privately to the manager and excluded from all shared documents.
[COMMAND_EXECUTION]: The skill interacts with external services exclusively through standardized MCP tools (google-workspace, claude_ai_Google_Calendar). It does not contain any patterns for arbitrary shell execution or dangerous system modifications.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests external data from 1:1 Docs and Plan Docs, it implements robust mitigations including strict parsing rules that ignore template blocks, evidence-tagging requirements, and multiple human-in-the-loop confirmation steps before any content is written to Google Drive or shared with other users.

planning-1on1-annual-digest