Single tool 'aip_security_status' has critical definition gaps. Input schema is empty object {} with no parameters documented. Description is present (58 chars) but minimal. No output schema documented. Tool appears to be a read-only status query but lacks the structured guidance LLMs need for proper invocation. The tool is registered via api.registerTool() in plugin/src/index.ts but definition quality is severely limited.

• Rename tool to 'get_aip_security_status' to follow verb_noun convention and clarify it retrieves current configuration.
• Expand description to 100-150 chars: 'Retrieve current AIP Guard security configuration: number of configured trust keys, unsigned skill policy (allow/warn/block), and audit log destination. Use this to verify security posture before executing tools.'
• Define explicit output schema with typed fields: { trust_keys_count: number, unsigned_policy: enum[allow|warn|block], audit_log_destination: string, configured: boolean }.
• Add parameter documentation even if empty: document that the tool takes no input parameters and explain why (it reads server-side config).
• Consider exposing additional tools for LLM interaction: 'verify_aip_token' (test token validity), 'list_audit_log_entries' (query recent decisions), 'check_tool_scope' (verify if a token grants access to a specific tool).