feat: image_edit tool — prompt-guided img2img editing with OpenAI Codex

acp_adapter/tools.py

@@ -50,6 +50,7 @@
     "delegate_task": "execute",
     "vision_analyze": "read",
     "image_generate": "execute",
+    "image_edit": "execute",
     "text_to_speech": "execute",
     # Thinking / meta
     "_thinking": "think",
@@ -65,7 +66,7 @@
     "skill_view", "skills_list", "skill_manage", "web_search", "web_extract",
     "browser_navigate", "browser_click", "browser_type", "browser_press", "browser_scroll",
     "browser_back", "browser_snapshot", "browser_console", "browser_get_images", "browser_vision",
-    "vision_analyze", "image_generate", "text_to_speech",
+    "vision_analyze", "image_generate", "image_edit", "text_to_speech",
     # Schedulers / platform integrations
     "cronjob", "send_message", "clarify", "discord", "discord_admin",
     "ha_list_entities", "ha_get_state", "ha_list_services", "ha_call_service",
@@ -173,6 +174,9 @@ def build_tool_title(tool_name: str, args: Dict[str, Any]) -> str:
     if tool_name == "image_generate":
         prompt = str(args.get("prompt") or args.get("description") or "").strip()
         return f"generate image: {prompt[:50]}" if prompt else "generate image"
+    if tool_name == "image_edit":
+        prompt = str(args.get("prompt") or args.get("instruction") or "").strip()
+        return f"edit image: {prompt[:50]}" if prompt else "edit image"
     if tool_name == "cronjob":
         action = str(args.get("action") or "manage").strip() or "manage"
         job_id = str(args.get("job_id") or args.get("id") or "").strip()
@@ -894,6 +898,7 @@ def _build_polished_completion_content(
         "browser_get_images": lambda: _format_browser_result(tool_name, result, function_args),
         "vision_analyze": lambda: _format_media_or_cron_result(tool_name, result),
         "image_generate": lambda: _format_media_or_cron_result(tool_name, result),
+        "image_edit": lambda: _format_media_or_cron_result(tool_name, result),
         "cronjob": lambda: _format_media_or_cron_result(tool_name, result),
     }.get(tool_name)
     if formatter is None and tool_name in _POLISHED_TOOLS:

agent/display.py

@@ -183,7 +183,7 @@ def build_tool_preview(tool_name: str, args: dict, max_len: int | None = None) -
         "read_file": "path", "write_file": "path", "patch": "path",
         "search_files": "pattern", "browser_navigate": "url",
         "browser_click": "ref", "browser_type": "text",
-        "image_generate": "prompt", "text_to_speech": "text",
+        "image_generate": "prompt", "image_edit": "prompt", "text_to_speech": "text",
         "vision_analyze": "question", "mixture_of_agents": "user_prompt",
         "skill_view": "name", "skills_list": "category",
         "cronjob": "action",
@@ -999,6 +999,8 @@ def _wrap(line: str) -> str:
         return _wrap(f"┊ 📚 skill     {_trunc(args.get('name', ''), 30)}  {dur}")
     if tool_name == "image_generate":
         return _wrap(f"┊ 🎨 create    {_trunc(args.get('prompt', ''), 35)}  {dur}")
+    if tool_name == "image_edit":
+        return _wrap(f"┊ 🖌️ edit      {_trunc(args.get('prompt', ''), 35)}  {dur}")
     if tool_name == "text_to_speech":
         return _wrap(f"┊ 🔊 speak     {_trunc(args.get('text', ''), 30)}  {dur}")
     if tool_name == "vision_analyze":

agent/image_gen_provider.py

@@ -142,6 +142,38 @@ def generate(
         should ignore unknown keys.
         """
 
+    def supports_edit(self) -> bool:
+        """Return True when this backend supports reference-image editing.
+
+        Providers that support image-to-image editing via :meth:`edit` should
+        override this to return ``True``.  The default returns ``False`` so
+        existing providers stay source-compatible.
+        """
+        return False
+
+    def edit(
+        self,
+        prompt: str,
+        image: str,
+        aspect_ratio: str = DEFAULT_ASPECT_RATIO,
+        **kwargs: Any,
+    ) -> Dict[str, Any]:
+        """Edit an existing image using a text instruction.
+
+        ``image`` is an HTTP(S) URL, data URL, or absolute local file path.
+        Providers that support image-to-image should override this method.
+        The default returns a uniform unsupported error so callers don't need
+        to probe :meth:`supports_edit` first.
+        """
+        aspect = resolve_aspect_ratio(aspect_ratio)
+        return error_response(
+            error=f"Image editing is not supported by provider '{self.name}'",
+            error_type="unsupported_operation",
+            provider=self.name,
+            prompt=prompt or "",
+            aspect_ratio=aspect,
+        )
+
 
 # ---------------------------------------------------------------------------
 # Helpers

agent/prompt_builder.py

@@ -1254,6 +1254,7 @@ def build_nous_subscription_prompt(valid_tool_names: "set[str] | None" = None) -
         "browser_get_images",
         "browser_vision",
         "image_generate",
+        "image_edit",
         "text_to_speech",
         "terminal",
         "process",

agent/transports/hermes_tools_mcp_server.py

@@ -80,6 +80,7 @@
     "browser_vision",
     "vision_analyze",
     "image_generate",
+    "image_edit",
     "skill_view",
     "skills_list",
     "text_to_speech",

hermes_cli/tools_config.py

@@ -59,7 +59,7 @@
     ("code_execution",  "⚡ Code Execution",            "execute_code"),
     ("vision",          "👁️  Vision / Image Analysis",  "vision_analyze"),
     ("video",           "🎬 Video Analysis",            "video_analyze (requires video-capable model)"),
-    ("image_gen",       "🎨 Image Generation",          "image_generate"),
+    ("image_gen",       "🎨 Image Generation / Editing", "image_generate / image_edit"),
     ("video_gen",       "🎬 Video Generation",          "video_generate (text-to-video + image-to-video)"),
     ("x_search",        "🐦 X (Twitter) Search",        "x_search (requires xAI OAuth or XAI_API_KEY)"),
     ("moa",             "🧠 Mixture of Agents",         "mixture_of_agents"),

model_tools.py

@@ -222,7 +222,7 @@ def _run_in_worker():
     "terminal_tools": ["terminal"],
     "vision_tools": ["vision_analyze"],
     "moa_tools": ["mixture_of_agents"],
-    "image_tools": ["image_generate"],
+    "image_tools": ["image_generate", "image_edit"],
     "skills_tools": ["skills_list", "skill_view", "skill_manage"],
     "browser_tools": [
         "browser_navigate", "browser_snapshot", "browser_click",