The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script, registry-events.py, located within the repository's scripts/ directory to process consent events and check suppression status.
[COMMAND_EXECUTION]: The instructions provide a shell snippet to dynamically resolve the project root using git rev-parse --show-toplevel or the CLAUDE_PLUGIN_ROOT environment variable.
[DATA_EXPOSURE]: The skill manages a local data stream in memory/events/consent.ndjson and a projection cache in memory/projections/consent-suppressions.json. However, it includes strict instructions to only store pseudonymous subject IDs and opaque proof references, specifically forbidding the storage of email addresses, phone numbers, or names.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted evidence and external event data. It mitigates risk by requiring host-capability authorization for mutations, enforcing schema validation via the registry script, and explicitly stating that external evidence cannot self-declare its own lawful basis.

consent-registry