The Agent Skills Directory

[SAFE]: Extensive analysis of the skill instructions and associated reference files confirms the behavior is consistent with its stated purpose of automated test generation. No malicious patterns or bypass techniques were detected.\n- [COMMAND_EXECUTION]: The skill uses the Bash tool for project environment discovery, utilizing standard utilities like jq, fd, and ripgrep. These operations are restricted to identifying dependencies and file paths within the current project and do not involve unauthorized system modifications.\n- [DATA_EXFILTRATION]: While the skill reads local project configuration files and source code to gather context for the LLM, these operations are performed locally. No instructions were found that attempt to transmit sensitive project data to external or untrusted endpoints.\n- [REMOTE_CODE_EXECUTION]: Skill instructions guide the agent to suggest or run generated tests through standard package managers and testing frameworks. These are routine development actions and do not involve fetching or executing untrusted remote scripts.

testgen