The Agent Skills Directory

[SAFE]: The skill performs a security and maturity audit of the local development environment ('harness'). It accesses local configuration files such as ~/.claude.json and .mcp.json to analyze installed plugins and MCP servers. This access is necessary for the audit and the data is used to generate a local report.
[COMMAND_EXECUTION]: Uses Bash to create temporary directories in /tmp for caching data and to open the generated HTML report using the system's default browser. These commands are standard for this type of utility and are executed locally.
[SAFE]: The skill uses sub-agents to process metadata. These agents are explicitly instructed not to read prompt text from sessions, ensuring that the content of user conversations is not processed during the pattern analysis.

check-harness