MCP Tools
Context Intelligence

apply_finding_fix

Apply Marcora's stored recommended fix for one or more pending content_grounding or Context Intelligence health_audit findings by UUID. Review suggested_fix with the user first — writes to Reference Library, Brand Foundation, or checked content documents.

Parameters

NameTypeRequiredDefaultDescription
team_idintegerYes—Team id to act in (from initialize instructions or get_current_user_info). Required on external MCP; optional in bound sessions.
finding_idsarray of stringYes—UUIDs of content_grounding or health_audit findings (from check_content_grounding, get_grounding_result, list_ci_findings, or get_ci_finding).
context_item_overridesobjectNo—Optional map of finding_id to target context_item_id UUID to redirect a library fix.

Input Schema

{
  "type": "object",
  "required": [
    "team_id",
    "finding_ids"
  ],
  "properties": {
    "team_id": {
      "type": "integer",
      "minimum": 1,
      "description": "Team id to act in (from initialize instructions or get_current_user_info). Required on external MCP; optional in bound sessions."
    },
    "finding_ids": {
      "type": "array",
      "items": {
        "type": "string",
        "format": "uuid"
      },
      "minItems": 1,
      "description": "UUIDs of content_grounding or health_audit findings (from check_content_grounding, get_grounding_result, list_ci_findings, or get_ci_finding)."
    },
    "context_item_overrides": {
      "type": "object",
      "description": "Optional map of finding_id to target context_item_id UUID to redirect a library fix.",
      "additionalProperties": {
        "type": "string",
        "format": "uuid"
      }
    }
  }
}

Output Schema

{
  "type": "object",
  "required": [
    "requested",
    "queued",
    "skipped",
    "jobs",
    "errors"
  ],
  "properties": {
    "jobs": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "finding_ids",
          "target_type",
          "target_id",
          "generation_id",
          "status"
        ],
        "properties": {
          "status": {
            "type": "string"
          },
          "target_id": {
            "type": "string"
          },
          "finding_ids": {
            "type": "array",
            "items": {
              "type": "string",
              "format": "uuid"
            }
          },
          "target_type": {
            "enum": [
              "context_item",
              "content",
              "core_context"
            ],
            "type": "string"
          },
          "generation_id": {
            "type": "string",
            "format": "uuid"
          }
        }
      }
    },
    "team": {
      "type": "object",
      "required": [
        "id",
        "name"
      ],
      "properties": {
        "id": {
          "type": "integer"
        },
        "name": {
          "type": "string"
        }
      },
      "description": "Team this action applied to."
    },
    "errors": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "finding_id",
          "error"
        ],
        "properties": {
          "error": {
            "type": "string"
          },
          "finding_id": {
            "type": "string",
            "format": "uuid"
          }
        }
      }
    },
    "queued": {
      "type": "integer",
      "description": "Number of findings queued."
    },
    "skipped": {
      "type": "integer",
      "description": "Number of findings skipped."
    },
    "requested": {
      "type": "integer",
      "description": "Number of finding IDs requested."
    }
  }
}

Instructions

Applies Marcora's stored recommended fix for one or more pending findings (content_grounding or Context Intelligence health_audit) by UUID. If the user already reviewed suggested_fix and supplies a finding_id, call this directly; otherwise read suggested_fix (from check_content_grounding, get_grounding_result, or list_ci_findings) and confirm with the user first — this writes to their Reference Library, Brand Foundation, or checked content document. Requires an Admin or Creator role and uses Studio Credits.

Routing & targets

  • content_grounding findings: update the target context item or document (context_item_id: null or conflict_scope: "company_material" revises the document, never your facts; a company gap offers no add-to-library). Pass context_item_overrides only to redirect a library gap/fix item (every key must also be in finding_ids).
  • health_audit findings: update Reference Library context items or Brand Foundation (target_type: "core_context").
  • web_freshness / impact_scan findings: cannot be auto-applied — they return in errors[] pointing to update_ci_finding_status.

Asynchronous completion & grouping

Returns immediately with requested, queued, skipped, jobs[], errors[], and team. Poll each distinct generation_id in jobs[] via get_generation_status. Fixes whose target is the same document or context item share one generation_id (Brand Foundation fixes each keep their own run). On the get_generation_status result, check content.document_updated (true if actually changed, false if already covered) and report it honestly.

Partial success is normal: check both jobs[] and errors[].

Parameters

Parameter Type Required Description
team_id integer Yes* Team ID to act in (required on external MCP; optional in bound sessions)
finding_ids array of string Yes UUIDs of pending content_grounding or health_audit findings to apply
context_item_overrides object No Map of finding_id to context_item_id (or null) to redirect a library fix. Every key must appear in finding_ids

* team_id is required on external MCP.

Output

Field Type Description
requested integer Number of finding IDs requested
queued integer Number of findings queued
skipped integer Number of findings skipped (see errors[])
jobs array Queued job records (finding_ids[], target_type [context_item | content | core_context], target_id, generation_id, status)
errors array Per-finding errors (finding_id, error)
team object Team this action applied to (id, name)

Example prompts

  • "Apply the fix for that pricing conflict"
  • "Apply all three of those health audit recommendations"