RACI Matrix Validator

Overview

You send the tasks, the roles and the letters assigned between them; we return whether the matrix satisfies the rules you enabled, and one entry per finding with the task or role it concerns. The two rules everybody agrees on — exactly one accountable per task, at least one responsible per task — are switches you control, so the verdict reflects your policy and not an opinion of ours. Findings we have no authority over come back as warnings instead: a role that holds nothing, a task nobody is assigned to, the same letter merged in twice. Responsible and accountable on the same role is normal practice and allowed by default; any other combination a role ends up holding on one task says two contradictory things and is reported once, naming the whole set, so the report does not depend on the order you listed the assignments in. Set a ceiling on consulted roles per task or accountable tasks per role and we will flag the overload as well. Nothing is stored, and role and task names are accepted but never echoed.

RACI matrices fail in a handful of predictable ways — two people accountable, nobody responsible, one person accountable for everything, a role left in from the last reorganisation — and they fail quietly, because a spreadsheet does not check itself. Making the rules explicit switches means the check can run in a form, in a pipeline, or in a review, and give the same answer each time.

When to use it

Scope

Non-goals

• Does not judge whether the right people hold the right letters
• Does not invent an owner for a task that has none
• Does not store, publish or notify anything
• Does not rank tasks or roles by importance
• Does not echo the names you attach to tasks and roles

Properties

Capabilities

Cost & billing

Flat rate per invocation

File handling

How we handle your data

Input schema

Output schema

API contract

curl -sS -X POST "$BASE/v1/execute" -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"skillKey":"business.raci_validate","input":{"matrix":{"tasks":[{"id":"spec"}],"roles":[{"id":"pm"},{"id":"dev"}],"assignments":[{"taskId":"spec","roleId":"pm","letter":"A"},{"taskId":"spec","roleId":"dev","letter":"R"}]}}}'

import { Squillify } from "@squillify/sdk";

const sq = new Squillify({ apiKey: process.env.SQUILLIFY_API_KEY });

const { output } = await sq.execute("business.raci_validate", {"matrix":{"roles":[{"id":"pm"},{"id":"dev"},{"id":"qa"}],"tasks":[{"id":"spec"},{"id":"build"}],"assignments":[{"taskId":"spec","roleId":"pm","letter":"A"},{"taskId":"spec","roleId":"dev","letter":"R"},{"taskId":"spec","roleId":"qa","letter":"C"},{"taskId":"build","roleId":"pm","letter":"A"},{"taskId":"build","roleId":"dev","letter":"R"}]}});

Errors

Other limits

Limitations

• Does not judge whether the right people hold the right letters
• Does not invent an owner for a task that has none
• Does not store, publish or notify anything
• Does not rank tasks or roles by importance
• Does not echo the names you attach to tasks and roles
• Deciding who should be accountable — that is a management call
• Storing or publishing the matrix
• Matrices with letters beyond R, A, C and I: extend the vocabulary elsewhere first

How to call

POST /v1/execute (requires authentication)

{
  "skillKey": "business.raci_validate",
  "input": {
    "matrix": {
      "roles": [
        {
          "id": "pm"
        },
        {
          "id": "dev"
        },
        {
          "id": "qa"
        }
      ],
      "tasks": [
        {
          "id": "spec"
        },
        {
          "id": "build"
        }
      ],
      "assignments": [
        {
          "taskId": "spec",
          "roleId": "pm",
          "letter": "A"
        },
        {
          "taskId": "spec",
          "roleId": "dev",
          "letter": "R"
        },
        {
          "taskId": "spec",
          "roleId": "qa",
          "letter": "C"
        },
        {
          "taskId": "build",
          "roleId": "pm",
          "letter": "A"
        },
        {
          "taskId": "build",
          "roleId": "dev",
          "letter": "R"
        }
      ]
    }
  }
}

Examples

{
  "matrix": {
    "roles": [
      {
        "id": "pm"
      },
      {
        "id": "dev"
      },
      {
        "id": "qa"
      }
    ],
    "tasks": [
      {
        "id": "spec"
      },
      {
        "id": "build"
      }
    ],
    "assignments": [
      {
        "taskId": "spec",
        "roleId": "pm",
        "letter": "A"
      },
      {
        "taskId": "spec",
        "roleId": "dev",
        "letter": "R"
      },
      {
        "taskId": "spec",
        "roleId": "qa",
        "letter": "C"
      },
      {
        "taskId": "build",
        "roleId": "pm",
        "letter": "A"
      },
      {
        "taskId": "build",
        "roleId": "dev",
        "letter": "R"
      }
    ]
  }
}

{
  "valid": true,
  "summary": {
    "taskCount": 2,
    "roleCount": 3,
    "assignmentCount": 5,
    "issueCount": 0,
    "bySeverity": {},
    "truncated": false,
    "tasksWithoutAccountable": 0,
    "tasksWithoutResponsible": 0,
    "rolesWithoutAssignment": 0
  },
  "rules": {
    "requireSingleAccountable": true,
    "requireResponsible": true,
    "allowCombined": true,
    "maxIssues": 1000
  },
  "issues": []
}

{
  "matrix": {
    "roles": [
      {
        "id": "pm"
      },
      {
        "id": "dev"
      },
      {
        "id": "ops"
      }
    ],
    "tasks": [
      {
        "id": "spec"
      },
      {
        "id": "deploy"
      }
    ],
    "assignments": [
      {
        "taskId": "spec",
        "roleId": "pm",
        "letter": "A"
      },
      {
        "taskId": "spec",
        "roleId": "dev",
        "letter": "A"
      },
      {
        "taskId": "spec",
        "roleId": "dev",
        "letter": "R"
      },
      {
        "taskId": "deploy",
        "roleId": "ops",
        "letter": "C"
      }
    ]
  }
}

{
  "valid": false,
  "summary": {
    "taskCount": 2,
    "roleCount": 3,
    "assignmentCount": 4,
    "issueCount": 3,
    "bySeverity": {
      "ERROR": 3
    },
    "truncated": false,
    "tasksWithoutAccountable": 1,
    "tasksWithoutResponsible": 1,
    "rolesWithoutAssignment": 0
  },
  "rules": {
    "requireSingleAccountable": true,
    "requireResponsible": true,
    "allowCombined": true,
    "maxIssues": 1000
  },
  "issues": [
    {
      "code": "MULTIPLE_ACCOUNTABLE",
      "severity": "ERROR",
      "taskId": "spec",
      "detail": "2 roles are accountable for this task. Accountability that is shared is accountability nobody holds: keep one A."
    },
    {
      "code": "NO_ACCOUNTABLE",
      "severity": "ERROR",
      "taskId": "deploy",
      "detail": "No role is accountable for this task. Assign exactly one A."
    },
    {
      "code": "NO_RESPONSIBLE",
      "severity": "ERROR",
      "taskId": "deploy",
      "detail": "No role is responsible for this task. Assign at least one R."
    }
  ]
}

Recommended use cases

• Gate a project-setup form until every task has exactly one owner
• Audit an inherited responsibility matrix before a handover
• Find the roles that own too much, before the person holding them tells you
• Keep one definition of a valid RACI across several teams' tooling

Avoid for

• Deciding who should be accountable — that is a management call
• Storing or publishing the matrix
• Matrices with letters beyond R, A, C and I: extend the vocabulary elsewhere first

Related skills

Composable with

Safety & privacy

• Runs without calling any external service.
• Your payloads are never stored permanently.
• Telemetry captures metadata only — never your payload contents.
• We never train models on your payloads.
• No human reviews your payloads.

Tags