model-eol 0.9.0 → 0.9.1

package/bot/bot.mjs

@@ -67,6 +67,12 @@ const PLAN_SCHEMA = 'model-eol.plan/0.1'
 const BOT_SCHEMA = 'model-eol.bot/0.1'
 const EVAL_FAILURE_CHANNEL = 'configured-eval'
 const EVAL_ARTIFACT_MAX_BYTES = 64 * 1024 * 1024
+const EVAL_RESULT_OVERHEAD_BYTES = 4096
+
+export const evalArtifactBudget = (maxReportBytes, groupCount) => Math.min(
+  EVAL_ARTIFACT_MAX_BYTES,
+  Math.max(1024 * 1024, EVAL_RESULT_OVERHEAD_BYTES + groupCount * (6 * maxReportBytes + EVAL_RESULT_OVERHEAD_BYTES)),
+)
 const ISSUE_REASONS = new Set([
   'not-direct-api',
   'no-replacement',
@@ -223,7 +229,7 @@ const readJsonArtifact = (file, maxBytes, label) => {
   const absolute = asAbsolute(file)
   const artifact = readReportCapped(absolute, maxBytes + 1)
   if (artifact.missing) throw new Error(`${label} is missing`)
-  if (Buffer.byteLength(artifact.report) > maxBytes) throw new Error(`${label} exceeds ${maxBytes} bytes`)
+  if (artifact.truncated || Buffer.byteLength(artifact.report) > maxBytes) throw new Error(`${label} exceeds ${maxBytes} bytes`)
   try {
     return JSON.parse(artifact.report)
   } catch (error) {
@@ -233,7 +239,7 @@ const readJsonArtifact = (file, maxBytes, label) => {
 
 const readEvalResults = ({ file, config, commandOverride, plan, groups, baseHead }) => {
   const settings = effectiveEval(config, commandOverride)
-  const maxBytes = Math.min(EVAL_ARTIFACT_MAX_BYTES, Math.max(1024 * 1024, groups.length * (settings.max_report_bytes + 2048)))
+  const maxBytes = evalArtifactBudget(settings.max_report_bytes, groups.length)
   const artifact = readJsonArtifact(file, maxBytes, 'eval results artifact')
   if (artifact?.schema !== EVAL_SCHEMA) throw new Error(`refusing eval results schema ${artifact?.schema ?? 'missing'}; expected ${EVAL_SCHEMA}`)
   try {
@@ -712,12 +718,12 @@ const evalWorkspaceDrift = (root, planFiles, expectedHashes, expectedHead) => {
   return null
 }
 
-const writeJsonAtomic = (file, value) => {
+const writeTextAtomic = (file, text) => {
   const absolute = asAbsolute(file)
   fs.mkdirSync(path.dirname(absolute), { recursive: true })
   const temporary = `${absolute}.${process.pid}.${sha256(`${Date.now()}-${absolute}`).slice(0, 8)}.tmp`
   try {
-    fs.writeFileSync(temporary, `${JSON.stringify(value, null, 2)}\n`, { mode: 0o600 })
+    fs.writeFileSync(temporary, text, { mode: 0o600 })
     fs.renameSync(temporary, absolute)
   } finally {
     fs.rmSync(temporary, { force: true })
@@ -843,7 +849,19 @@ export const evaluatePlan = async ({
       configured: Boolean(settings.command),
       results,
     }
-    writeJsonAtomic(outputFile, artifact)
+    const budget = evalArtifactBudget(settings.max_report_bytes, groups.length)
+    let encoded
+    try {
+      encoded = `${JSON.stringify(artifact, null, 2)}\n`
+    } catch (error) {
+      if (!(error instanceof RangeError)) throw error
+      throw new Error(`refusing eval: eval results artifact is too large to encode, over its ${budget}-byte budget; lower eval.max_report_bytes`)
+    }
+    const encodedBytes = Buffer.byteLength(encoded)
+    if (encodedBytes > budget) {
+      throw new Error(`refusing eval: eval results artifact would be ${encodedBytes} bytes, over its ${budget}-byte budget; lower eval.max_report_bytes`)
+    }
+    writeTextAtomic(outputFile, encoded)
     return { artifact, plan, config: { ...config, eval: settings }, degraded: Boolean(feedSet.degraded) }
   } finally {
     feedSet.cleanup()

package/bot/lib/eval.mjs

@@ -2,21 +2,36 @@ import fs from 'node:fs'
 import { spawnSync } from 'node:child_process'
 import { markdownFenceText } from './common.mjs'
 
-const truncateReport = (bytes, limit, truncated) => {
-  if (!truncated) return bytes.toString('utf8')
-  const marker = Buffer.from('\n[model-eol: report truncated]', 'utf8')
-  if (limit <= marker.length) return marker.subarray(0, limit).toString('utf8')
-  return Buffer.concat([bytes.subarray(0, limit - marker.length), marker]).subarray(0, limit).toString('utf8')
+const TRUNCATION_MARKER = '\n[model-eol: report truncated]'
+
+const fitUtf8 = (text, limit) => {
+  const bytes = Buffer.from(text, 'utf8')
+  if (bytes.length <= limit) return text
+  let end = Math.max(0, limit)
+  while (end > 0 && (bytes[end] & 0xc0) === 0x80) end--
+  return bytes.subarray(0, end).toString('utf8')
+}
+
+const truncateReport = (text, limit) => {
+  const markerBytes = Buffer.byteLength(TRUNCATION_MARKER)
+  if (limit <= markerBytes) return fitUtf8(TRUNCATION_MARKER, limit)
+  return fitUtf8(text, limit - markerBytes) + TRUNCATION_MARKER
+}
+
+const byteCap = limit => {
+  if (!Number.isFinite(limit)) throw new Error(`report byte limit must be finite, got ${limit}`)
+  return Math.max(0, Math.floor(limit))
 }
 
 export const capReport = (report, limit) => {
   if (report === null || report === undefined) return null
-  const bytes = Buffer.from(String(report), 'utf8')
-  return truncateReport(bytes, Math.max(0, limit), bytes.length > limit)
+  const cap = byteCap(limit)
+  const text = Buffer.from(String(report), 'utf8').toString('utf8')
+  return Buffer.byteLength(text) > cap ? truncateReport(text, cap) : text
 }
 
 export const readReportCapped = (file, limit) => {
-  const cap = Math.max(0, limit)
+  const cap = byteCap(limit)
   let lstat
   try {
     lstat = fs.lstatSync(file)
@@ -41,10 +56,9 @@ export const readReportCapped = (file, limit) => {
   try {
     const buffer = Buffer.alloc(cap + 1)
     const bytesRead = fs.readSync(fd, buffer, 0, buffer.length, 0)
-    return {
-      missing: false,
-      report: truncateReport(buffer.subarray(0, Math.min(bytesRead, cap)), cap, bytesRead > cap),
-    }
+    const text = buffer.subarray(0, Math.min(bytesRead, cap)).toString('utf8')
+    const truncated = bytesRead > cap || Buffer.byteLength(text) > cap
+    return { missing: false, truncated, report: truncated ? truncateReport(text, cap) : text }
   } finally {
     fs.closeSync(fd)
   }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "model-eol",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "AI model deprecation feeds with a zero-dependency checker, inventory CLI, and migration bot.",
   "type": "module",
   "bin": {