12 tools. 8 can modify or destroy data without limits.
2 destructive tools with no built-in limits. Policy required.
Last updated:
Destructive tools (delete_card, delete_dashboard) permanently delete resources. There is no undo. An agent calling these in a retry loop causes irreversible damage.
Write operations (create_card, create_dashboard, update_card) modify state. Without rate limits, an agent can make hundreds of changes in seconds — faster than any human can review or revert.
Execute tools (execute_card, execute_query) trigger processes with side effects. Builds, notifications, workflows — all fired without throttling.
delete_card:
rules:
- action: deny Destructive tools should never be available to autonomous agents without human approval.
create_card:
rules:
- rate_limit: 30/hour Prevents bulk unintended modifications from agents caught in loops.
get_dashboard_cards:
rules:
- rate_limit: 60/minute Controls API costs and prevents retry loops from exhausting upstream rate limits.
Yes. The Metabase Integration Server server exposes 2 destructive tools including delete_card, delete_dashboard. These permanently remove resources with no undo. Intercept blocks destructive tools by default so they never reach the upstream server.
The Metabase Integration Server server has 4 write tools including create_card, create_dashboard, update_card. Set rate limits in your policy file -- for example, rate_limit: 10/hour prevents an agent from making more than 10 modifications per hour. Intercept enforces this at the transport layer.
12 tools across 4 categories: Destructive, Execute, Read, Write. 4 are read-only. 8 can modify, create, or delete data.
One line change. Instead of running the Metabase Integration Server server directly, prefix it with Intercept: intercept -c imlewc-metabase-server.yaml -- npx -y @imlewc/metabase-server. Download a pre-built policy from policylayer.com/policies/imlewc-metabase-server and adjust the limits to match your use case.
Starter policies available for each. Same risk classification, same one-command setup.
Deterministic policy on every MCP tool call. Per-identity grants. Full audit log.