diff --git a/CHANGELOG.md b/CHANGELOG.md
index d815dd4be..11abf7827 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,35 @@
 
 All notable changes to GBrain will be documented in this file.
 
+## [0.23.1] - 2026-04-30
+
+**Dream cycle can't eat its own output. Verdict model is configurable.**
+
+Two fixes for the dream-cycle synthesize phase shipped in v0.23.0.
+
+### Self-consumption guard (built-in, zero config)
+
+The transcript discovery layer now auto-skips any file whose first 2000 characters contain dream output slug prefixes (`wiki/personal/reflections/`, `wiki/originals/ideas/`, `wiki/personal/patterns/`, `dream-cycle-summaries/`). This prevents the dream cycle from synthesizing its own output if transcripts ever include dream-generated pages — an infinite recursion bug that would compound with each overnight cycle.
+
+The guard is structural and always-on. User-configured `exclude_patterns` are additive on top. The 2000-char head scan keeps the cost negligible (no full-content scan).
+
+### Configurable verdict model
+
+The significance verdict model (the cheap "is this worth processing?" filter) is now configurable via `dream.synthesize.verdict_model`. Default remains `claude-haiku-4-5-20251001`. Set it if you want a stronger model for the triage pass:
+
+```bash
+gbrain config set dream.synthesize.verdict_model claude-sonnet-4-6
+```
+
+### Itemized changes
+
+#### Fixed
+- `src/core/cycle/transcript-discovery.ts`: built-in `isDreamOutput()` guard checks first 2000 chars for dream output slug prefixes. Applied in both `discoverTranscripts()` and `readSingleTranscript()`. Runs before user-configured exclude patterns.
+- `src/core/cycle/synthesize.ts`: `judgeSignificance()` now accepts a `verdictModel` parameter (default `claude-haiku-4-5-20251001`). Loaded from `dream.synthesize.verdict_model` config key via `loadSynthConfig()`.
+
+#### Tests
+- 3 new test cases in `test/cycle-synthesize.test.ts` under `self-consumption guard`: discovery skips dream output files, single-transcript returns null for dream output, deep-buried slugs don't false-positive (performance guard checks head only).
+
 ## [0.23.0] - 2026-04-26
 
 **`gbrain dream` now actually dreams. Conversation transcripts become reflections, originals, and 25-year patterns ... overnight.**
diff --git a/VERSION b/VERSION
index ca222b7cf..610e28725 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-0.23.0
+0.23.1
diff --git a/package.json b/package.json
index 19c0c3492..bcfbb0443 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "gbrain",
-  "version": "0.23.0",
+  "version": "0.23.1",
   "description": "Postgres-native personal knowledge brain with hybrid RAG search",
   "type": "module",
   "main": "src/core/index.ts",
diff --git a/src/core/cycle/synthesize.ts b/src/core/cycle/synthesize.ts
index 96cd40e21..c6c5e8feb 100644
--- a/src/core/cycle/synthesize.ts
+++ b/src/core/cycle/synthesize.ts
@@ -120,7 +120,7 @@ export async function runPhaseSynthesize(
         verdicts.push({ filePath: t.filePath, worth: false, reasons: ['no ANTHROPIC_API_KEY for significance judge'], cached: false });
         continue;
       }
-      const verdict = await judgeSignificance(haiku, t);
+      const verdict = await judgeSignificance(haiku, t, config.verdictModel);
       await engine.putDreamVerdict(t.filePath, t.contentHash, verdict);
       verdicts.push({ filePath: t.filePath, worth: verdict.worth_processing, reasons: verdict.reasons, cached: false });
       if (verdict.worth_processing) worthProcessing.push(t);
@@ -248,6 +248,7 @@ interface SynthConfig {
   minChars: number;
   excludePatterns: string[];
   model: string;
+  verdictModel: string;
   cooldownHours: number;
 }
 
@@ -258,6 +259,7 @@ async function loadSynthConfig(engine: BrainEngine): Promise<SynthConfig> {
   const minCharsStr = await engine.getConfig('dream.synthesize.min_chars');
   const excludeStr = await engine.getConfig('dream.synthesize.exclude_patterns');
   const model = (await engine.getConfig('dream.synthesize.model')) || 'claude-sonnet-4-6';
+  const verdictModel = (await engine.getConfig('dream.synthesize.verdict_model')) || 'claude-haiku-4-5-20251001';
   const cooldownHoursStr = await engine.getConfig('dream.synthesize.cooldown_hours');
 
   let excludePatterns: string[] = ['medical', 'therapy'];
@@ -275,6 +277,7 @@ async function loadSynthConfig(engine: BrainEngine): Promise<SynthConfig> {
     minChars: minCharsStr ? Math.max(0, parseInt(minCharsStr, 10) || 2000) : 2000,
     excludePatterns,
     model,
+    verdictModel,
     cooldownHours: cooldownHoursStr ? Math.max(0, parseInt(cooldownHoursStr, 10) || 12) : 12,
   };
 }
@@ -336,6 +339,7 @@ interface VerdictResult {
 async function judgeSignificance(
   client: JudgeClient,
   t: DiscoveredTranscript,
+  verdictModel = 'claude-haiku-4-5-20251001',
 ): Promise<VerdictResult> {
   // Truncate the transcript at 8K chars for cost control. Haiku's verdict
   // doesn't need the full body; the opening + closing sections are usually
@@ -362,7 +366,7 @@ Respond as JSON: {"worth_processing": <bool>, "reasons": ["<short>", "<short>"]}
 Two reasons max, one phrase each.`;
 
   const msg = await client.create({
-    model: 'claude-haiku-4-5-20251001',
+    model: verdictModel,
     max_tokens: 200,
     system: sys,
     messages: [{ role: 'user', content: `Transcript ${t.basename}:\n\n${trimmed}` }],
diff --git a/src/core/cycle/transcript-discovery.ts b/src/core/cycle/transcript-discovery.ts
index 31ece6264..d1ee0cd81 100644
--- a/src/core/cycle/transcript-discovery.ts
+++ b/src/core/cycle/transcript-discovery.ts
@@ -46,6 +46,27 @@ export interface DiscoverOpts {
 const DATE_RE = /^(\d{4}-\d{2}-\d{2})/;
 const WORD_BOUNDARY_HEURISTIC = /^[a-zA-Z][a-zA-Z0-9_-]*$/;
 
+/**
+ * Built-in self-consumption guard: slug prefixes that indicate dream output.
+ * If a transcript contains any of these paths, it's the dream cycle's own
+ * output being fed back in — skip it to prevent infinite recursion.
+ * This is a hard-coded safety net; user-configured exclude_patterns are
+ * additive on top of this.
+ */
+const DREAM_OUTPUT_SLUGS = [
+  'wiki/personal/reflections/',
+  'wiki/originals/ideas/',
+  'wiki/personal/patterns/',
+  'dream-cycle-summaries/',
+];
+
+function isDreamOutput(content: string): boolean {
+  // Check the first 2000 chars (frontmatter + opening) for dream output markers.
+  // Full-content scan is wasteful; dream output always self-identifies early.
+  const head = content.slice(0, 2000);
+  return DREAM_OUTPUT_SLUGS.some(slug => head.includes(slug));
+}
+
 /**
  * Auto-wrap bare-word patterns in `\b<word>\b`. Power users can pass full
  * regex (e.g. `^therapy:`) which we honor verbatim. Heuristic: any input
@@ -141,6 +162,7 @@ export function discoverTranscripts(opts: DiscoverOpts): DiscoveredTranscript[]
         continue;
       }
       if (content.length < minChars) continue;
+      if (isDreamOutput(content)) continue;
       if (matchesAnyExclude(content, excludeRes)) continue;
 
       results.push({
@@ -175,6 +197,7 @@ export function readSingleTranscript(
     throw new Error(`could not read transcript at ${filePath}: ${msg}`);
   }
   if (content.length < minChars) return null;
+  if (isDreamOutput(content)) return null;
   if (matchesAnyExclude(content, excludeRes)) return null;
   const baseName = basename(filePath, '.txt');
   const dateMatch = DATE_RE.exec(baseName);
diff --git a/test/cycle-synthesize.test.ts b/test/cycle-synthesize.test.ts
index ed1ba294c..0aa7c1ded 100644
--- a/test/cycle-synthesize.test.ts
+++ b/test/cycle-synthesize.test.ts
@@ -189,3 +189,36 @@ describe('readSingleTranscript', () => {
     expect(t!.inferredDate).toBeNull();
   });
 });
+
+describe('self-consumption guard', () => {
+  test('discoverTranscripts skips files containing dream output slug prefixes', () => {
+    makeTranscript('2026-04-25-real.txt', 'This is a real conversation transcript. ' + 'x'.repeat(3000));
+    makeTranscript('2026-04-25-dream.txt',
+      '---\ntitle: Reflection\n---\n# Some reflection\nwiki/personal/reflections/2026-04-25-foo-abc123\n' + 'x'.repeat(3000));
+    makeTranscript('2026-04-25-original.txt',
+      '---\ntitle: Idea\n---\n# Some idea\nwiki/originals/ideas/2026-04-25-bar-def456\n' + 'x'.repeat(3000));
+    makeTranscript('2026-04-25-pattern.txt',
+      '---\ntype: pattern\n---\nwiki/personal/patterns/shame-cycle\n' + 'x'.repeat(3000));
+    makeTranscript('2026-04-25-summary.txt',
+      'dream-cycle-summaries/2026-04-25 generated this page\n' + 'x'.repeat(3000));
+
+    const results = discoverTranscripts({ corpusDir: tmpDir, minChars: 1000 });
+    expect(results).toHaveLength(1);
+    expect(results[0].basename).toBe('2026-04-25-real');
+  });
+
+  test('readSingleTranscript returns null for dream output', () => {
+    const path = makeTranscript('2026-04-25-dream.txt',
+      'wiki/personal/reflections/2026-04-25-thing-abc123\n' + 'x'.repeat(3000));
+    const result = readSingleTranscript(path, { minChars: 1000 });
+    expect(result).toBeNull();
+  });
+
+  test('self-consumption guard checks only first 2000 chars (performance)', () => {
+    // Dream slug buried deep in the file should NOT trigger the guard
+    const content = 'x'.repeat(3000) + 'wiki/personal/reflections/2026-04-25-buried-abc123';
+    const path = makeTranscript('2026-04-25-deep.txt', content);
+    const result = readSingleTranscript(path, { minChars: 1000 });
+    expect(result).not.toBeNull();
+  });
+});