The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the Bash tool for executing shell commands such as find, stat, wc, tail, grep, and sqlite3. These commands are used to inspect file metadata and retrieve status counts from the OrchestKit coordination database located in the user's local state directory (~/.local/state/orchestkit/).
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it reads and processes the content of various telemetry and log files. These files store historical agent decisions and outputs which could contain malicious instructions if the logged data originated from untrusted sources.
Ingestion points: Reads files in the project's .claude/ subdirectories (telemetry, logs, state, and feedback).
Boundary markers: No specific delimiters or instructions are provided to the agent to treat log content as untrusted data rather than instructions.
Capability inventory: The skill uses Bash, Read, Grep, and Glob tools to analyze and report on file contents.
Sanitization: No sanitization or validation of log content is performed before the data is summarized for the health report.

telemetry-inspect