Skip to main content
This page lists the MCP tools the Oplane server exposes to your AI assistant. To connect your assistant and see the workflows that use these tools, see Threat model with MCP. You don’t need to call these tools directly. They run automatically as part of the MCP workflows.

How read tools work

Read tools come in pairs:
  • list_* tools find records. By default they return matching IDs. Ask for a summary and they also return a summary table of the matches.
  • get_* tools fetch records by ID. Each call accepts up to 100 IDs, so one call can fetch a full page of list_* results.
Your assistant typically calls a list_* tool to narrow down what it needs, then a get_* tool to fetch the full records. Each read covers one organisation. If you belong to more than one, the tool returns your organisations so your assistant can pick one. Read tools have their own per-user rate limit of 300 calls per minute. This is separate from the budget the other tools share. To ask your assistant questions about your security data, such as open findings by severity or time to resolve, see Insights with MCP.

Tools

  • Statuses and severity, for the values update_implementation_state and update_security_requirement_severity accept.
  • Applications, to call these tools from a CI pipeline with an API key.
  • Policies, for how policies created with these tools are matched and applied.