get_finding_documentation

Reference for the get_finding_documentation tool of the SparkForensics MCP server.

Detection and tuning reference documentation for one finding type, independent of any run: fetch it once per type (not once per finding) and cache it.

Parameters:

• type: string (required): a finding type value, e.g. "skew", or a detector-level type such as "broadcastSizing", which returns the documentation of the finding types it emits

Example call:

{ "name": "get_finding_documentation", "arguments": { "type": "skew" } }

Example response:

{
  "type": "skew",
  "name": "Task Skew",
  "detectionDoc": {
    "tag": "SKEW",
    "title": "Task skew",
    "content": "### `SKEW`: Task skew {#skew}\n\nA small number of tasks..."
  },
  "tuningDoc": {
    "anchor": "#bottleneck-skew",
    "title": "Task Skew",
    "content": "# Task Skew\n\n..."
  }
}

tuningDoc is null when the finding type has no vendored tuning-doc page. Two types have none: configAudit (its four audited properties each have their own anchor rather than one shared page, so no single anchor resolves) and incompleteRun (no upstream tuning page covers this signal at all). A type whose section lives on a general chapter rather than a bottleneck page (autoscalingChurn on Cluster Tuning, cacheUtilization on Memory Management) returns that whole chapter, with anchor naming the section.