JSON output

Every command supports --format json for machine-readable output. The schemas are versioned: version is 1, and changes within v1 are additive only, so parsers can rely on the fields below. Keys are camelCase, optional fields are omitted when absent, and file paths are relative to root.

`roe --format json`

The combined run — a bare roe, or roe check — prints a single document rather than three concatenated ones, so stdout stays valid JSON:

{
  "version": 1,
  "root": "/path/to/solution",
  "deadCode": { "version": 1, "…": "…" },
  "dupes": { "version": 1, "…": "…" },
  "health": { "version": 1, "…": "…" }
}

Field	Description
`version`	Schema version of the combined document, `1`.
`root`	The analysis root path.
`deadCode`	The `roe dead-code` report, unchanged.
`dupes`	The `roe dupes` report, unchanged. `mode` is always `exact` — the combined run takes no `--mode` flag.
`health`	The `roe health` report, unchanged. `hotspots` is always empty — hotspots need `roe health --hotspots`.

Each nested value is byte-for-byte the schema the individual command emits, including its own version and root, so anything already written against roe dead-code --format json can consume deadCode as-is.

`roe dead-code --format json`

{
  "version": 1,
  "root": "/path/to/solution",
  "summary": {
    "projects": 3,
    "filesScanned": 214,
    "symbols": 2610,
    "unusedTypes": 0,
    "unusedMembers": 1,
    "unusedFiles": 1,
    "elapsedMs": 74
  },
  "notes": [],
  "findings": [
    {
      "kind": "unused-member",
      "symbolKind": "method",
      "name": "App.Billing.InvoiceService.RecalculateAll",
      "project": "App",
      "file": "src/App/Billing/InvoiceService.cs",
      "line": 88,
      "column": 17,
      "visibility": "private"
    }
  ]
}

Field	Description
`version`	Schema version, `1`.
`root`	The analysis root path.
`summary`	Scan totals: `projects`, `filesScanned`, `symbols`, `unusedTypes`, `unusedMembers`, `unusedFiles`, `elapsedMs`.
`notes`	Human-readable notes about the scan, if any.
`findings[].kind`	`unused-type`, `unused-member`, or `unused-file` — the same names used by inline suppressions.
`findings[].symbolKind`	The kind of symbol (e.g. `method`). Optional.
`findings[].name`	Fully-qualified symbol name, or the file path for a dead file.
`findings[].project`	The containing project. Optional.
`findings[].file`, `line`, `column`	Location of the finding; dead-file findings are pinned at `1:1`.
`findings[].visibility`	The symbol’s visibility (e.g. `private`). Optional.

`roe dupes --format json`

{
  "version": 1,
  "root": "/path/to/solution",
  "mode": "exact",
  "summary": {
    "projects": 1,
    "filesScanned": 214,
    "groups": 1,
    "duplicatedLines": 48,
    "elapsedMs": 61
  },
  "groups": [
    {
      "tokenCount": 105,
      "lineCount": 24,
      "occurrences": [
        {
          "file": "src/App/Billing/PaymentService.cs",
          "startLine": 40,
          "startColumn": 5,
          "endLine": 63,
          "endColumn": 6
        },
        {
          "file": "src/App/Shipping/ShippingService.cs",
          "startLine": 38,
          "startColumn": 5,
          "endLine": 61,
          "endColumn": 6
        }
      ]
    }
  ]
}

Field	Description
`version`	Schema version, `1`.
`root`	The analysis root path.
`mode`	The matching mode used: `exact` or `semantic`.
`summary`	Scan totals: `projects`, `filesScanned`, `groups`, `duplicatedLines`, `elapsedMs`.
`groups[].tokenCount`	Length of the duplicated block in tokens.
`groups[].lineCount`	Length of the duplicated block in lines.
`groups[].occurrences[]`	Every location of the block: `file`, `startLine`, `startColumn`, `endLine`, `endColumn`.

`roe health --format json`

{
  "version": 1,
  "root": "/path/to/solution",
  "summary": {
    "projects": 1,
    "filesScanned": 214,
    "symbols": 2610,
    "highComplexity": 1,
    "highCognitiveComplexity": 0,
    "longMethods": 0,
    "tooManyParameters": 1,
    "largeFiles": 0,
    "largeTypes": 1,
    "circularDependencies": 1,
    "elapsedMs": 8,
    "excluded": {
      "testProjects": ["App.Tests"],
      "ignoredFiles": 4
    }
  },
  "findings": [
    {
      "kind": "large-type",
      "name": "App.Billing.InvoiceService",
      "project": "App",
      "file": "src/App/Billing/InvoiceService.cs",
      "line": 3,
      "column": 14,
      "metric": 34,
      "threshold": 20,
      "breakdown": {
        "methods": 15,
        "properties": 19,
        "fields": 0,
        "events": 0
      }
    },
    {
      "kind": "too-many-parameters",
      "name": "App.Billing.InvoiceService.Reconcile",
      "project": "App",
      "file": "src/App/Billing/InvoiceService.cs",
      "line": 42,
      "column": 17,
      "metric": 8,
      "threshold": 5,
      "parameters": {
        "required": 8,
        "optional": 2,
        "out": 1
      }
    }
  ],
  "cycles": [
    {
      "path": [
        {
          "name": "App.Orders.Order",
          "project": "App",
          "file": "src/App/Orders/Order.cs",
          "line": 3,
          "column": 14
        },
        {
          "name": "App.Orders.Invoice",
          "project": "App",
          "file": "src/App/Orders/Invoice.cs",
          "line": 8,
          "column": 14
        }
      ],
      "others": []
    }
  ],
  "hotspots": []
}

Field	Description
`version`	Schema version, `1`.
`root`	The analysis root path.
`summary`	Scan totals: `projects`, `filesScanned`, `symbols`, one count per check (`highComplexity`, `highCognitiveComplexity`, `longMethods`, `tooManyParameters`, `largeFiles`, `largeTypes`, `circularDependencies`), and `elapsedMs`. `commitsWalked` is present only when `--hotspots` was passed. The three scan totals count what was eligible to be reported, so they narrow under `--exclude-tests` and the config’s `ignore` globs (top-level and `health.ignore`).
`summary.baselined`	How many findings and cycles a baseline hid — they are counted nowhere else in this document. Present only when a baseline was in force, so `0` means “the baseline is fully ratcheted” and absent means “no baseline was used”.
`summary.excluded`	What was ruled out before any check ran: `testProjects` (an array of project names, present under `--exclude-tests`) and `ignoredFiles` (a count). Omitted entirely when nothing was excluded.
`findings[].kind`	`high-complexity`, `high-cognitive-complexity`, `long-method`, `too-many-parameters`, `large-file`, or `large-type` — the same names used by inline suppressions.
`findings[].name`	Fully-qualified symbol name, or the file path for a large file.
`findings[].project`	The containing project. Optional.
`findings[].file`, `line`, `column`	Location of the finding; large-file findings are pinned at `1:1`.
`findings[].metric`	The measured value — the complexity score, line count, member count, or, for `too-many-parameters`, the required parameter count rather than the declared total.
`findings[].threshold`	The threshold it exceeded. `metric / threshold` is the severity the human report sorts by.
`findings[].breakdown`	Member composition — `methods`, `properties`, `fields`, `events`. Present on `large-type` findings only.
`findings[].parameters`	Signature composition — `required`, `optional`, `out`. Present on `too-many-parameters` findings only. `required` always equals `metric`; the three sum to the declared parameter count.
`cycles[].path`	The types on the cycle, each referencing the next and the last referencing the first. Each entry has `name`, `project`, `file`, `line`, `column`.
`cycles[].others`	Types in the same tangle that aren’t on `path`. Empty for a simple cycle.
`hotspots[]`	Present only with `--hotspots`: `file`, `project`, `score`, `weightedCommits`, `cyclomatic`, `lines`, `complexityDensity`. Empty otherwise.

One finding is emitted per tripped check, so a single method that’s both too complex and too long appears twice. The human report groups those into one entry, and --sort and --limit apply to it alone — JSON is never reordered or truncated.