fix(codex): default gpt-5.2-codex to xhigh and sync docs

Author: ndycode

.gitignore

@@ -15,4 +15,9 @@ Heap-*.heapsnapshot
 .codex-cache
 .sisyphus/
 coverage/
-winston/
+winston/
+.tmp-plugin-test/
+.tmp-release-*.md
+task_plan.md
+findings.md
+progress.md

CHANGELOG.md

@@ -15,6 +15,8 @@ all notable changes to this project. dates are ISO format (YYYY-MM-DD).
 - **request correlation**: each upstream fetch now sets a correlation id, reuses `CODEX_THREAD_ID`/`prompt_cache_key` when available, and clears scope after each request.
 - **plan-mode tool gating**: `request_user_input` is automatically stripped from tool definitions when collaboration mode is Default (kept in Plan mode).
 - **safety prompt hardening**: bridge/remap prompts now explicitly block destructive git commands unless the user asks for them.
+- **gpt-5.2-codex default effort**: default reasoning now prefers `xhigh` when no explicit effort/variant is provided.
+- **gitignore hygiene**: local planning/release scratch artifacts are now ignored to keep working trees clean.
 
 ### fixed
 

README.md

@@ -130,7 +130,7 @@ opencode run "Hello" --model=openai/gpt-5.2 --variant=medium
 | Model | Variants | Notes |
 |-------|----------|-------|
 | `gpt-5.2` | none, low, medium, high, xhigh | Latest GPT-5.2 with reasoning levels |
-| `gpt-5.2-codex` | low, medium, high, xhigh | GPT-5.2 Codex for code generation |
+| `gpt-5.2-codex` | low, medium, high, xhigh | GPT-5.2 Codex for code generation (default: xhigh) |
 | `gpt-5.1-codex-max` | low, medium, high, xhigh | Maximum context Codex |
 | `gpt-5.1-codex` | low, medium, high | Standard Codex |
 | `gpt-5.1-codex-mini` | medium, high | Lightweight Codex |
@@ -185,6 +185,10 @@ Add this to your `~/.config/opencode/opencode.json`:
             "medium": { "reasoningEffort": "medium" },
             "high": { "reasoningEffort": "high" },
             "xhigh": { "reasoningEffort": "xhigh" }
+          },
+          "options": {
+            "reasoningEffort": "xhigh",
+            "reasoningSummary": "detailed"
           }
         },
         "gpt-5.1-codex-max": {

config/opencode-modern.json

@@ -95,6 +95,11 @@
               "reasoningSummary": "detailed",
               "textVerbosity": "medium"
             }
+          },
+          "options": {
+            "reasoningEffort": "xhigh",
+            "reasoningSummary": "detailed",
+            "textVerbosity": "medium"
           }
         },
         "gpt-5.1-codex-max": {

docs/configuration.md

@@ -35,7 +35,7 @@ controls how much thinking the model does.
 | model | supported values |
 |-------|------------------|
 | `gpt-5.2` | none, low, medium, high, xhigh |
-| `gpt-5.2-codex` | low, medium, high, xhigh |
+| `gpt-5.2-codex` | low, medium, high, xhigh (default: xhigh) |
 | `gpt-5.1-codex-max` | low, medium, high, xhigh |
 | `gpt-5.1-codex` | low, medium, high |
 | `gpt-5.1-codex-mini` | medium, high |

lib/request/request-transformer.ts

@@ -317,11 +317,14 @@ export function getReasoningConfig(
 	// - Codex models (including GPT-5.2 Codex) do NOT support "none"
 	const supportsNone = isGpt52General || isGpt51General;
 
-	// Default based on model type (Codex CLI defaults)
+	// Default based on model type (Codex CLI defaults + plugin opinionated tuning)
 	// Note: OpenAI docs say gpt-5.1 defaults to "none", but we default to "medium"
-	// for better coding assistance unless user explicitly requests "none"
+	// for better coding assistance unless user explicitly requests "none".
+	// GPT-5.2 Codex defaults to xhigh in this plugin for stronger coding output.
 	const defaultEffort: ReasoningConfig["effort"] = isCodexMini
 		? "medium"
+		: isGpt52Codex
+			? "xhigh"
 		: supportsXhigh
 			? "high"
 			: isLightweight

test/request-transformer.test.ts

@@ -896,14 +896,14 @@ describe('Request Transformer Module', () => {
 			expect(result.reasoning?.effort).toBe('high');
 		});
 
-		it('should default gpt-5.2-codex to high effort', async () => {
+		it('should default gpt-5.2-codex to xhigh effort', async () => {
 			const body: RequestBody = {
 				model: 'gpt-5.2-codex',
 				input: [],
 			};
 			const result = await transformRequestBody(body, codexInstructions);
 			expect(result.model).toBe('gpt-5.2-codex');
-			expect(result.reasoning?.effort).toBe('high');
+			expect(result.reasoning?.effort).toBe('xhigh');
 		});
 
 		it('should preserve xhigh for codex-max when requested', async () => {