diff --git a/.agents/skills/infographic-agent/README.md b/.agents/skills/infographic-agent/README.md
index 2c1167c..f9e4ad9 100644
--- a/.agents/skills/infographic-agent/README.md
+++ b/.agents/skills/infographic-agent/README.md
@@ -24,7 +24,7 @@ npx infographic-agent "Top 5 programming languages in 2026"
 
 ## 📋 Prerequisites
 
-- **Python 3.8+**
+- **Python 3.9+**
 - **Gemini API Key**: Get a free key at [Google AI Studio](https://aistudio.google.com/apikey). Set it as:
   ```bash
   export GEMINI_API_KEY="your-key"
diff --git a/.agents/skills/infographic-agent/SKILL.md b/.agents/skills/infographic-agent/SKILL.md
index a6051fb..665e0d0 100644
--- a/.agents/skills/infographic-agent/SKILL.md
+++ b/.agents/skills/infographic-agent/SKILL.md
@@ -6,7 +6,7 @@ description: >
   then gemini-3.1-flash-lite-image renders it into a PNG. No browser, Playwright, or Chromium dependencies —
   the only requirement is Google's GenAI SDK. Fully portable to any agent CLI environment.
 metadata:
-  version: "3.2.0"
+  version: "3.2.2"
   author: "Infographic Agent contributors"
 ---
 
diff --git a/.agents/skills/infographic-agent/bin/infographic-agent.js b/.agents/skills/infographic-agent/bin/infographic-agent.js
index 95bf0ce..a43f0ac 100755
--- a/.agents/skills/infographic-agent/bin/infographic-agent.js
+++ b/.agents/skills/infographic-agent/bin/infographic-agent.js
@@ -51,7 +51,7 @@ function findPython() {
       if (match) {
         const major = parseInt(match[1], 10);
         const minor = parseInt(match[2], 10);
-        if (major >= 3 && minor >= 8) return candidate;
+        if (major >= 3 && minor >= 9) return candidate;
       }
     }
   }
@@ -120,7 +120,7 @@ Examples:
 
 if (!python) {
   err(
-    "Python 3.8+ is required but was not found on your PATH.\n\n" +
+    "Python 3.9+ is required but was not found on your PATH.\n\n" +
     "  Install Python from https://www.python.org/downloads/ and make sure\n" +
     "  it is on your PATH, then re-run:\n\n" +
     "    npx infographic-agent ..."
@@ -132,7 +132,7 @@ if (!python) {
 
 if (doInstall) {
   print("Installing Python dependencies (google-genai, pillow)...");
-  const result = run(python, ["-m", "pip", "install", "--upgrade", "google-genai", "pillow"]);
+  const result = run(python, ["-m", "pip", "install", "--upgrade", "google-genai>=1.47.0", "pillow"]);
   if (result.status !== 0) process.exit(result.status);
 
   print(`
diff --git a/.agents/skills/infographic-agent/package.json b/.agents/skills/infographic-agent/package.json
index 8a1fb50..ff587c1 100644
--- a/.agents/skills/infographic-agent/package.json
+++ b/.agents/skills/infographic-agent/package.json
@@ -1,6 +1,6 @@
 {
   "name": "infographic-agent",
-  "version": "3.2.0",
+  "version": "3.2.2",
   "description": "Generate professional infographic PNGs from text with Gemini — a research agent (gemini-3.5-flash) prepares and validates the prompt, then gemini-3.1-flash-lite-image renders it. No browser/Playwright deps.",
   "license": "MIT",
   "author": "Ryan Baumann",
diff --git a/.agents/skills/infographic-agent/portable_infographic.py b/.agents/skills/infographic-agent/portable_infographic.py
index bc33b6f..2b67230 100755
--- a/.agents/skills/infographic-agent/portable_infographic.py
+++ b/.agents/skills/infographic-agent/portable_infographic.py
@@ -637,9 +637,14 @@ def _normalize_to_png(data: bytes, mime: str):
         return data, m
 
 
+def _config_fields(config_type):
+    """Return the declared fields across supported Pydantic SDK versions."""
+    return getattr(config_type, "model_fields", None) or getattr(config_type, "__fields__", {})
+
+
 def _image_config(aspect: str, resolution: str):
-    """Use newer quality controls only when the installed SDK supports them."""
-    fields = getattr(types.ImageConfig, "model_fields", {})
+    """Request image size only when the installed SDK exposes that field."""
+    fields = _config_fields(types.ImageConfig)
     kwargs = {"aspect_ratio": aspect}
     if "image_size" in fields:
         kwargs["image_size"] = resolution
@@ -649,7 +654,8 @@ def _image_config(aspect: str, resolution: str):
 
 
 def _thinking_config():
-    fields = getattr(types.ThinkingConfig, "model_fields", {})
+    """Use high thinking on current SDKs and dynamic thinking on older SDKs."""
+    fields = _config_fields(types.ThinkingConfig)
     if "thinking_level" in fields:
         return types.ThinkingConfig(thinking_level="HIGH", include_thoughts=True)
     return types.ThinkingConfig(thinking_budget=-1, include_thoughts=True)
diff --git a/.agents/skills/infographic-agent/requirements.txt b/.agents/skills/infographic-agent/requirements.txt
index bb94f99..704b837 100644
--- a/.agents/skills/infographic-agent/requirements.txt
+++ b/.agents/skills/infographic-agent/requirements.txt
@@ -1,4 +1,4 @@
 # Runtime dependencies. Install with:  pip install -r requirements.txt
 # (or:  pip install google-genai pillow)
-google-genai>=1.0.0   # Gemini SDK — the two-agent research + image pipeline
+google-genai>=1.47.0  # Gemini SDK — includes the typed image/thinking config surface tested here
 pillow>=10.0.0        # transcodes the model's output to lossless PNG (crisp text)
diff --git a/.agents/skills/infographic-agent/test_portable_infographic.py b/.agents/skills/infographic-agent/test_portable_infographic.py
new file mode 100644
index 0000000..464ccb9
--- /dev/null
+++ b/.agents/skills/infographic-agent/test_portable_infographic.py
@@ -0,0 +1,62 @@
+import importlib.util
+import unittest
+from pathlib import Path
+from types import SimpleNamespace
+from unittest.mock import patch
+
+
+SCRIPT_PATH = Path(__file__).with_name("portable_infographic.py")
+SPEC = importlib.util.spec_from_file_location("portable_infographic", SCRIPT_PATH)
+portable = importlib.util.module_from_spec(SPEC)
+SPEC.loader.exec_module(portable)
+
+
+def config_type(*field_names):
+    class Config:
+        model_fields = {name: object() for name in field_names}
+
+        def __init__(self, **kwargs):
+            self.kwargs = kwargs
+
+    return Config
+
+
+class SDKConfigCompatibilityTest(unittest.TestCase):
+    def test_current_sdk_requests_resolution_and_high_thinking(self):
+        fake_types = SimpleNamespace(
+            ImageConfig=config_type("aspect_ratio", "image_size"),
+            ThinkingConfig=config_type("thinking_level", "include_thoughts"),
+        )
+
+        with patch.object(portable, "types", fake_types):
+            image = portable._image_config("16:9", "1K")
+            thinking = portable._thinking_config()
+
+        self.assertEqual(image.kwargs, {"aspect_ratio": "16:9", "image_size": "1K"})
+        self.assertEqual(
+            thinking.kwargs,
+            {"thinking_level": "HIGH", "include_thoughts": True},
+        )
+
+    def test_sdk_1_47_falls_back_to_supported_fields(self):
+        fake_types = SimpleNamespace(
+            ImageConfig=config_type("aspect_ratio"),
+            ThinkingConfig=config_type("thinking_budget", "include_thoughts"),
+        )
+
+        with patch.object(portable, "types", fake_types), patch.object(portable, "warn") as warn:
+            image = portable._image_config("16:9", "1K")
+            thinking = portable._thinking_config()
+
+        self.assertEqual(image.kwargs, {"aspect_ratio": "16:9"})
+        self.assertEqual(
+            thinking.kwargs,
+            {"thinking_budget": -1, "include_thoughts": True},
+        )
+        warn.assert_called_once_with(
+            "Installed google-genai does not expose image_size; using the model's native resolution."
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()
diff --git a/CHANGELOG.md b/CHANGELOG.md
index b945ee1..cefc5bd 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,7 +7,7 @@
 - Replaced the oversized text-heavy hero with a 16:9 growth-loop diagram, added a separate 1200×627 social card, and embedded a purpose-built agent-evaluation loop in the article.
 - Added reusable front-matter aliases for future detail-page renames and documented the workflow in the portfolio content skill.
 - Hardened aliases against slashless duplicates, canonical-page collisions, private-route disclosure, stale canonicals, external targets, and bodyless entries.
-- Installed Infographic Agent 3.2.0 and documented the required three-asset visual standard for every future hosted essay.
+- Updated the vendored Infographic Agent skill to the merged 3.2.1 source, including Python 3.9 / `google-genai 1.47.0` compatibility and matching lock metadata, and documented the required three-asset visual standard for every future hosted essay.
 - Archived the exact final image prompts and settings, including Gemini 3.1 Flash Image at 1K with high thinking, so all three visuals can be reproduced.
 - Added a mandatory portfolio review workflow for every publishable content change: deterministic copy, claim, link, URL, image, metadata, and browser checks plus a bounded independent maker/checker loop.
 
diff --git a/skills-lock.json b/skills-lock.json
index 92a60c6..562373b 100644
--- a/skills-lock.json
+++ b/skills-lock.json
@@ -5,7 +5,7 @@
       "source": "ryanbaumann/infographic-agent",
       "sourceType": "github",
       "skillPath": "skill/infographic-agent/SKILL.md",
-      "computedHash": "ec4962103a11264951bc9d431af6a9f6721e1df8f84e4df1c98bb6a2e0a2a416"
+      "computedHash": "9f2d5eb57d968332ef3d0fecb1b7c320967c4c820cb786ef86c3d8442681d3c6"
     }
   }
 }