Switch default codex model from gpt-5.3-codex to gpt-5.2-codex (v1.12.2)

SihaoLiu · SihaoLiu · commit 0c32d617b6d4 · 2026-03-04T09:27:17.000-08:00
diff --git a/.claude-plugin/marketplace.json b/.claude-plugin/marketplace.json
@@ -8,7 +8,7 @@
       "name": "humanize",
       "source": "./",
       "description": "Humanize - An iterative development plugin that uses Codex to review Claude's work. Creates a feedback loop where Claude implements plans and Codex independently reviews progress, ensuring quality through continuous refinement.",
-      "version": "1.12.1"
+      "version": "1.12.2"
     }
   ]
 }
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
@@ -1,7 +1,7 @@
 {
   "name": "humanize",
   "description": "Humanize - An iterative development plugin that uses Codex to review Claude's work. Creates a feedback loop where Claude implements plans and Codex independently reviews progress, ensuring quality through continuous refinement.",
-  "version": "1.12.1",
+  "version": "1.12.2",
   "author": {
     "name": "humania-org"
   },
diff --git a/README.md b/README.md
@@ -1,6 +1,6 @@
 # Humanize
 
-**Current Version: 1.12.1**
+**Current Version: 1.12.2**
 
 > Derived from the [GAAC (GitHub-as-a-Context)](https://github.com/SihaoLiu/gaac) project.
 
@@ -151,7 +151,7 @@ OPTIONS:
   --plan-file <path>     Explicit plan file path (alternative to positional arg)
   --max <N>              Maximum iterations before auto-stop (default: 42)
   --codex-model <MODEL:EFFORT>
-                         Codex model and reasoning effort (default: gpt-5.3-codex:xhigh)
+                         Codex model and reasoning effort (default: gpt-5.2-codex:xhigh)
   --codex-timeout <SECONDS>
                          Timeout for each Codex review in seconds (default: 5400)
   --track-plan-file      Indicate plan file should be tracked in git (must be clean)
@@ -216,7 +216,7 @@ OPTIONS:
 
 OPTIONS:
   --codex-model <MODEL:EFFORT>
-                         Codex model and reasoning effort (default: gpt-5.3-codex:xhigh)
+                         Codex model and reasoning effort (default: gpt-5.2-codex:xhigh)
   --codex-timeout <SECONDS>
                          Timeout for the Codex query in seconds (default: 3600)
   -h, --help             Show help message
diff --git a/hooks/lib/loop-common.sh b/hooks/lib/loop-common.sh
@@ -46,12 +46,12 @@ readonly FIELD_AGENT_TEAMS="agent_teams"
 # Detect Claude Code plugin mode by checking if this script lives under
 # .claude/plugins/cache/ (CLAUDE_PLUGIN_ROOT is a config-time template
 # variable, not a runtime env var, so we cannot check it directly).
-#   - Claude Code plugin mode -> gpt-5.3-codex
+#   - Claude Code plugin mode -> gpt-5.2-codex
 #   - Skill mode (Codex/Kimi) -> gpt-5.2
 if [[ -z "${DEFAULT_CODEX_MODEL:-}" ]]; then
     _LOOP_COMMON_SELF="${BASH_SOURCE[0]:-$0}"
     if [[ "$_LOOP_COMMON_SELF" == */.claude/plugins/cache/* ]]; then
-        DEFAULT_CODEX_MODEL="gpt-5.3-codex"
+        DEFAULT_CODEX_MODEL="gpt-5.2-codex"
     else
         DEFAULT_CODEX_MODEL="gpt-5.2"
     fi
diff --git a/scripts/ask-codex.sh b/scripts/ask-codex.sh
@@ -55,7 +55,7 @@ USAGE:
 
 OPTIONS:
   --codex-model <MODEL:EFFORT>
-                       Codex model and reasoning effort (default: gpt-5.3-codex:xhigh)
+                       Codex model and reasoning effort (default: gpt-5.2-codex:xhigh)
   --codex-timeout <SECONDS>
                        Timeout for the Codex query in seconds (default: 3600)
   -h, --help           Show this help message
@@ -68,7 +68,7 @@ DESCRIPTION:
 
 EXAMPLES:
   /humanize:ask-codex How should I structure the authentication module?
-  /humanize:ask-codex --codex-model gpt-5.3-codex:high What are the performance bottlenecks?
+  /humanize:ask-codex --codex-model gpt-5.2-codex:high What are the performance bottlenecks?
   /humanize:ask-codex --codex-timeout 300 Review the error handling in src/api/
 
 ENVIRONMENT:
diff --git a/skills/ask-codex/SKILL.md b/skills/ask-codex/SKILL.md
@@ -35,4 +35,4 @@ Execute the ask-codex script with the user's arguments:
 ## Notes
 
 - The response is saved to `.humanize/skill/<timestamp>/output.md` for reference
-- Default model is `gpt-5.3-codex:xhigh` with a 3600-second timeout
+- Default model is `gpt-5.2-codex:xhigh` with a 3600-second timeout
diff --git a/tests/robustness/test-state-file-robustness.sh b/tests/robustness/test-state-file-robustness.sh
@@ -36,7 +36,7 @@ cat > "$TEST_DIR/state.md" << 'EOF'
 ---
 current_round: 5
 max_iterations: 10
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 codex_timeout: 5400
 push_every_round: false
@@ -79,7 +79,7 @@ max_iterations: 20
 extra_field: some_value
 another_extra: 12345
 custom_metadata: true
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 codex_timeout: 5400
 ---
@@ -404,7 +404,7 @@ cat > "$TEST_DIR/state-full-review.md" << 'EOF'
 current_round: 3
 max_iterations: 20
 full_review_round: 7
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 plan_file: plan.md
 plan_tracked: false
@@ -429,7 +429,7 @@ cat > "$TEST_DIR/state-no-full-review.md" << 'EOF'
 ---
 current_round: 2
 max_iterations: 15
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 plan_file: plan.md
 plan_tracked: false
@@ -455,7 +455,7 @@ cat > "$TEST_DIR/state-min-full-review.md" << 'EOF'
 current_round: 1
 max_iterations: 10
 full_review_round: 2
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 plan_file: plan.md
 plan_tracked: false
diff --git a/tests/test-agent-teams.sh b/tests/test-agent-teams.sh
@@ -441,7 +441,7 @@ GI_EOF
 ---
 current_round: $round
 max_iterations: 42
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 codex_timeout: 5400
 push_every_round: false
diff --git a/tests/test-finalize-phase.sh b/tests/test-finalize-phase.sh
@@ -191,7 +191,7 @@ setup_loop_dir() {
 ---
 current_round: $round
 max_iterations: $max_iter
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 codex_timeout: 5400
 push_every_round: false
diff --git a/tests/test-plan-file-validation.sh b/tests/test-plan-file-validation.sh
@@ -717,7 +717,7 @@ fi
 echo "Test 13: Reject codex effort with YAML-unsafe characters"
 rm -rf "$TEST_DIR/.humanize/rlcr" 2>/dev/null || true
 set +e
-RESULT=$("$PROJECT_ROOT/scripts/setup-rlcr-loop.sh" --codex-model "gpt-5.3-codex:high#comment" "plans/test-plan.md" 2>&1)
+RESULT=$("$PROJECT_ROOT/scripts/setup-rlcr-loop.sh" --codex-model "gpt-5.2-codex:high#comment" "plans/test-plan.md" 2>&1)
 EXIT_CODE=$?
 set -e
 if [[ $EXIT_CODE -ne 0 ]] && echo "$RESULT" | grep -q "invalid characters"; then
@@ -729,7 +729,7 @@ fi
 # Test 14: Accept valid codex model with dots and hyphens
 echo "Test 14: Accept valid codex model (alphanumeric, dots, hyphens)"
 set +e
-RESULT=$("$PROJECT_ROOT/scripts/setup-rlcr-loop.sh" --codex-model "gpt-5.3-codex:medium" "plans/test-plan.md" 2>&1)
+RESULT=$("$PROJECT_ROOT/scripts/setup-rlcr-loop.sh" --codex-model "gpt-5.2-codex:medium" "plans/test-plan.md" 2>&1)
 EXIT_CODE=$?
 set -e
 # Should not fail due to model/effort validation (may fail later for other reasons)
diff --git a/tests/test-session-id.sh b/tests/test-session-id.sh
@@ -120,7 +120,7 @@ cat > "$TEST_DIR/project/.humanize/rlcr/2026-01-01_00-00-00/state.md" << 'EOF'
 ---
 current_round: 0
 max_iterations: 10
-codex_model: gpt-5.3-codex
+codex_model: gpt-5.2-codex
 codex_effort: xhigh
 codex_timeout: 5400
 push_every_round: false
diff --git a/tests/test-skill-monitor.sh b/tests/test-skill-monitor.sh
@@ -58,7 +58,7 @@ setup_test_env() {
 create_skill_invocation() {
     local unique_id="$1"
     local status="$2"
-    local model="${3:-gpt-5.3-codex}"
+    local model="${3:-gpt-5.2-codex}"
     local effort="${4:-xhigh}"
     local duration="${5:-15s}"
     local question="${6:-How should I structure this?}"
@@ -140,7 +140,7 @@ echo "=== Skill Monitor: Single Invocation ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_21-02-35-12345-abc123" "success" "gpt-5.3-codex" "xhigh" "15s" "How should I structure the auth module?"
+create_skill_invocation "2026-02-19_21-02-35-12345-abc123" "success" "gpt-5.2-codex" "xhigh" "15s" "How should I structure the auth module?"
 
 output=$(_humanize_monitor_skill --once 2>&1) && rc=0 || rc=$?
 if [[ $rc -eq 0 ]]; then
@@ -167,7 +167,7 @@ else
     fail "Should show success status" "got: $output"
 fi
 
-if echo "$output" | grep -q "gpt-5.3-codex"; then
+if echo "$output" | grep -q "gpt-5.2-codex"; then
     pass "Shows model name"
 else
     fail "Should show model" "got: $output"
@@ -199,10 +199,10 @@ echo "=== Skill Monitor: Multiple Invocations ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_20-00-00-111-aaa" "success" "gpt-5.3-codex" "high" "10s" "First question"
-create_skill_invocation "2026-02-19_20-30-00-222-bbb" "error" "gpt-5.3-codex" "xhigh" "5s" "Second question"
-create_skill_invocation "2026-02-19_21-00-00-333-ccc" "timeout" "gpt-5.3-codex" "xhigh" "3600s" "Third question"
-create_skill_invocation "2026-02-19_21-30-00-444-ddd" "success" "gpt-5.3-codex" "xhigh" "20s" "Latest question"
+create_skill_invocation "2026-02-19_20-00-00-111-aaa" "success" "gpt-5.2-codex" "high" "10s" "First question"
+create_skill_invocation "2026-02-19_20-30-00-222-bbb" "error" "gpt-5.2-codex" "xhigh" "5s" "Second question"
+create_skill_invocation "2026-02-19_21-00-00-333-ccc" "timeout" "gpt-5.2-codex" "xhigh" "3600s" "Third question"
+create_skill_invocation "2026-02-19_21-30-00-444-ddd" "success" "gpt-5.2-codex" "xhigh" "20s" "Latest question"
 
 output=$(_humanize_monitor_skill --once 2>&1) && rc=0 || rc=$?
 if echo "$output" | grep -q "Total Invocations: 4"; then
@@ -250,8 +250,8 @@ echo "=== Skill Monitor: Running Invocation ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_21-00-00-111-aaa" "success" "gpt-5.3-codex" "xhigh" "10s" "Completed question"
-create_skill_invocation "2026-02-19_21-30-00-222-bbb" "running" "gpt-5.3-codex" "high" "" "Running question"
+create_skill_invocation "2026-02-19_21-00-00-111-aaa" "success" "gpt-5.2-codex" "xhigh" "10s" "Completed question"
+create_skill_invocation "2026-02-19_21-30-00-222-bbb" "running" "gpt-5.2-codex" "high" "" "Running question"
 
 output=$(_humanize_monitor_skill --once 2>&1) && rc=0 || rc=$?
 if echo "$output" | grep -q "Running: 1"; then
@@ -274,9 +274,9 @@ echo "=== Skill Monitor: Recent Invocations List ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_20-00-00-111-aaa" "success" "gpt-5.3-codex" "xhigh" "10s" "Question one"
-create_skill_invocation "2026-02-19_20-30-00-222-bbb" "error" "gpt-5.3-codex" "xhigh" "5s" "Question two"
-create_skill_invocation "2026-02-19_21-00-00-333-ccc" "success" "gpt-5.3-codex" "xhigh" "20s" "Question three"
+create_skill_invocation "2026-02-19_20-00-00-111-aaa" "success" "gpt-5.2-codex" "xhigh" "10s" "Question one"
+create_skill_invocation "2026-02-19_20-30-00-222-bbb" "error" "gpt-5.2-codex" "xhigh" "5s" "Question two"
+create_skill_invocation "2026-02-19_21-00-00-333-ccc" "success" "gpt-5.2-codex" "xhigh" "20s" "Question three"
 
 output=$(_humanize_monitor_skill --once 2>&1) && rc=0 || rc=$?
 if echo "$output" | grep -q "Recent Invocations"; then
@@ -314,13 +314,13 @@ Additional context about the question.
 
 ## Configuration
 
-- Model: gpt-5.3-codex
+- Model: gpt-5.2-codex
 - Effort: xhigh
 - Timeout: 3600s
 EOF
 cat > "$local_dir/metadata.md" << 'EOF'
 ---
-model: gpt-5.3-codex
+model: gpt-5.2-codex
 effort: xhigh
 timeout: 3600
 exit_code: 0
@@ -353,7 +353,7 @@ echo "=== Skill Monitor: Empty Response ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_21-00-00-111-aaa" "empty_response" "gpt-5.3-codex" "xhigh" "30s" "Why is the sky blue?"
+create_skill_invocation "2026-02-19_21-00-00-111-aaa" "empty_response" "gpt-5.2-codex" "xhigh" "30s" "Why is the sky blue?"
 
 output=$(_humanize_monitor_skill --once 2>&1) && rc=0 || rc=$?
 if echo "$output" | grep -q "Empty: 1"; then
@@ -376,7 +376,7 @@ echo "=== Skill Monitor: Non-skill Dir Filtering ==="
 
 setup_test_env
 mkdir -p .humanize/skill
-create_skill_invocation "2026-02-19_21-00-00-111-aaa" "success" "gpt-5.3-codex" "xhigh" "10s" "Real question"
+create_skill_invocation "2026-02-19_21-00-00-111-aaa" "success" "gpt-5.2-codex" "xhigh" "10s" "Real question"
 # Create a non-matching directory
 mkdir -p ".humanize/skill/not-a-skill-dir"
 echo "junk" > ".humanize/skill/not-a-skill-dir/input.md"

Original file line number	Diff line number	Diff line change
`@@ -8,7 +8,7 @@`
`8`	`8`	`"name": "humanize",`
`9`	`9`	`"source": "./",`
`10`	`10`	`"description": "Humanize - An iterative development plugin that uses Codex to review Claude's work. Creates a feedback loop where Claude implements plans and Codex independently reviews progress, ensuring quality through continuous refinement.",`
`11`		`- "version": "1.12.1"`
	`11`	`+ "version": "1.12.2"`
`12`	`12`	`}`
`13`	`13`	`]`
`14`	`14`	`}`
Original file line number	Diff line number	Diff line change
`@@ -1,7 +1,7 @@`
`1`	`1`	`{`
`2`	`2`	`"name": "humanize",`
`3`	`3`	`"description": "Humanize - An iterative development plugin that uses Codex to review Claude's work. Creates a feedback loop where Claude implements plans and Codex independently reviews progress, ensuring quality through continuous refinement.",`
`4`		`- "version": "1.12.1",`
	`4`	`+ "version": "1.12.2",`
`5`	`5`	`"author": {`
`6`	`6`	`"name": "humania-org"`
`7`	`7`	`},`