@mseep/open-computer-use 1.0.0

package/openwebui/patches/fix_tool_loop_errors.py

@@ -0,0 +1,510 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: FSL-1.1-Apache-2.0
+# Copyright (c) 2025 Open Computer Use Contributors
+"""Unified patch: error handling for tool loop, code interpreter, SSE, and background tasks.
+
+Replaces 3 separate patches:
+- fix_debug_streaming_errors.py (TEMPORARY debug logging)
+- fix_transport_retry.py (transport error handling + non-streaming errors)
+- fix_tool_loop_error.py (budget error display + background tasks)
+
+Problems solved:
+1. Tool loop errors silently swallowed (log.debug + break) — user sees garbage
+2. Transport errors (aiohttp.ClientPayloadError) leave partial output
+3. Non-streaming errors (JSONResponse 400) silently break
+4. "Model not found" on budget exhaustion — cryptic message for user
+5. Background tasks (follow_ups/title/tags) crash on 404 after budget exceeded
+6. SSE parse errors logged at debug level only
+
+Applied at Docker build time. Works on ORIGINAL middleware.py (no dependencies).
+Target: Open WebUI 0.9.5 (output-based architecture, serialize_output, prior_output).
+
+Fail-loud: ANY sub-anchor miss triggers sys.exit(1) with stderr ERROR — refuses
+to ship a partially-patched middleware.py. Idempotent: re-run prints ALREADY PATCHED.
+"""
+
+import os
+import sys
+
+_PATCH_TARGET_OVERRIDE = os.environ.get("_PATCH_TARGET_OVERRIDE", "")
+MIDDLEWARE_PATH = _PATCH_TARGET_OVERRIDE or "/app/backend/open_webui/utils/middleware.py"
+
+PATCH_MARKER = "TOOL_LOOP_ERRORS_UNIFIED"   # legacy marker — retained
+NEW_PATCH_MARKER = "FIX_TOOL_LOOP_ERRORS"   # v0.9.2.0 marker (appears in injected comments)
+
+_BUDGET_MSG = (
+    "Model temporarily unavailable. "
+    "Request limit may be exceeded. "
+    "Try again later or choose another model."
+)
+_TRANSPORT_MSG = (
+    "Connection error while receiving model response. "
+    "Resend message to continue."
+)
+_ERROR_LABEL = "Error"
+
+# ============================================================
+# Mod 1: Tool loop — full error handling
+# ============================================================
+SEARCH_TOOL_LOOP = """\
+                    try:
+                        new_form_data = {
+                            **form_data,
+                            'model': model_id,
+                            'stream': True,
+                            'metadata': metadata,
+                        }
+
+                        if ENABLE_RESPONSES_API_STATEFUL and last_response_id:
+                            system_message = get_system_message(form_data['messages'])
+                            new_form_data['messages'] = (
+                                [system_message] if system_message else []
+                            ) + convert_output_to_messages(
+                                output, raw=True, reasoning_format=get_reasoning_format(model)
+                            )
+                            new_form_data['previous_response_id'] = last_response_id
+                        else:
+                            tool_messages = convert_output_to_messages(
+                                output, raw=True, reasoning_format=get_reasoning_format(model)
+                            )
+
+                            # Chat Completions providers don't support multimodal
+                            # tool messages.  Extract images into a user message.
+                            image_urls = []
+                            for message in tool_messages:
+                                if message.get('role') == 'tool' and isinstance(message.get('content'), list):
+                                    text_parts = []
+                                    for part in message['content']:
+                                        if part.get('type') == 'input_text':
+                                            text_parts.append(part.get('text', ''))
+                                        elif part.get('type') == 'input_image':
+                                            image_urls.append(part.get('image_url', ''))
+                                    message['content'] = ''.join(text_parts)
+
+                            new_form_data['messages'] = [
+                                *form_data['messages'],
+                                *tool_messages,
+                            ]
+
+                            if image_urls:
+                                new_form_data['messages'].append(
+                                    {
+                                        'role': 'user',
+                                        'content': [
+                                            {
+                                                'type': 'text',
+                                                'text': 'Here are the images from the tool results above. Please analyze them.',
+                                            },
+                                            *[{'type': 'image_url', 'image_url': {'url': url}} for url in image_urls],
+                                        ],
+                                    }
+                                )
+
+                        res = await generate_chat_completion(
+                            request,
+                            new_form_data,
+                            user,
+                            bypass_system_prompt=True,
+                        )
+
+                        if isinstance(res, StreamingResponse):
+                            # Save accumulated output and start fresh.
+                            # Responses API output_index values are relative
+                            # to the current response — a clean output list
+                            # keeps indices aligned. The display prefix
+                            # ensures the UI shows tool history during
+                            # streaming.
+                            prior_output = list(output)
+                            # Trim the trailing empty placeholder message
+                            # so it doesn't persist as a ghost item once
+                            # the new stream produces real content.
+                            if (
+                                prior_output
+                                and prior_output[-1].get('type') == 'message'
+                                and prior_output[-1].get('status') == 'in_progress'
+                            ):
+                                msg_parts = prior_output[-1].get('content', [])
+                                if not msg_parts or (len(msg_parts) == 1 and not msg_parts[0].get('text', '').strip()):
+                                    prior_output.pop()
+                            output = []
+                            await stream_body_handler(res, new_form_data)
+                            output[:0] = prior_output
+                            prior_output = []
+                        else:
+                            break
+                    except Exception as e:
+                        log.debug(e)
+                        break
+
+                if DETECT_CODE_INTERPRETER:"""
+
+REPLACE_TOOL_LOOP = (
+    "                    _saved_output = json.loads(json.dumps(output))  # TOOL_LOOP_ERRORS_UNIFIED: save for restore on error\n"
+    "                    try:\n"
+    "                        new_form_data = {\n"
+    "                            **form_data,\n"
+    "                            'model': model_id,\n"
+    "                            'stream': True,\n"
+    "                            'metadata': metadata,\n"
+    "                        }\n"
+    "\n"
+    "                        if ENABLE_RESPONSES_API_STATEFUL and last_response_id:\n"
+    "                            system_message = get_system_message(form_data['messages'])\n"
+    "                            new_form_data['messages'] = (\n"
+    "                                [system_message] if system_message else []\n"
+    "                            ) + convert_output_to_messages(\n"
+    "                                output, raw=True, reasoning_format=get_reasoning_format(model)\n"
+    "                            )\n"
+    "                            new_form_data['previous_response_id'] = last_response_id\n"
+    "                        else:\n"
+    "                            tool_messages = convert_output_to_messages(\n"
+    "                                output, raw=True, reasoning_format=get_reasoning_format(model)\n"
+    "                            )\n"
+    "\n"
+    "                            # Chat Completions providers don't support multimodal\n"
+    "                            # tool messages.  Extract images into a user message.\n"
+    "                            image_urls = []\n"
+    "                            for message in tool_messages:\n"
+    "                                if message.get('role') == 'tool' and isinstance(message.get('content'), list):\n"
+    "                                    text_parts = []\n"
+    "                                    for part in message['content']:\n"
+    "                                        if part.get('type') == 'input_text':\n"
+    "                                            text_parts.append(part.get('text', ''))\n"
+    "                                        elif part.get('type') == 'input_image':\n"
+    "                                            image_urls.append(part.get('image_url', ''))\n"
+    "                                    message['content'] = ''.join(text_parts)\n"
+    "\n"
+    "                            new_form_data['messages'] = [\n"
+    "                                *form_data['messages'],\n"
+    "                                *tool_messages,\n"
+    "                            ]\n"
+    "\n"
+    "                            if image_urls:\n"
+    "                                new_form_data['messages'].append(\n"
+    "                                    {\n"
+    "                                        'role': 'user',\n"
+    "                                        'content': [\n"
+    "                                            {\n"
+    "                                                'type': 'text',\n"
+    "                                                'text': 'Here are the images from the tool results above. Please analyze them.',\n"
+    "                                            },\n"
+    "                                            *[{'type': 'image_url', 'image_url': {'url': url}} for url in image_urls],\n"
+    "                                        ],\n"
+    "                                    }\n"
+    "                                )\n"
+    "\n"
+    "                        res = await generate_chat_completion(\n"
+    "                            request,\n"
+    "                            new_form_data,\n"
+    "                            user,\n"
+    "                            bypass_system_prompt=True,\n"
+    "                        )\n"
+    "\n"
+    "                        if isinstance(res, StreamingResponse):\n"
+    "                            # Save accumulated output and start fresh.\n"
+    "                            # Responses API output_index values are relative\n"
+    "                            # to the current response -- a clean output list\n"
+    "                            # keeps indices aligned. The display prefix\n"
+    "                            # ensures the UI shows tool history during\n"
+    "                            # streaming.\n"
+    "                            prior_output = list(output)\n"
+    "                            # Trim the trailing empty placeholder message\n"
+    "                            # so it doesn't persist as a ghost item once\n"
+    "                            # the new stream produces real content.\n"
+    "                            if (\n"
+    "                                prior_output\n"
+    "                                and prior_output[-1].get('type') == 'message'\n"
+    "                                and prior_output[-1].get('status') == 'in_progress'\n"
+    "                            ):\n"
+    "                                msg_parts = prior_output[-1].get('content', [])\n"
+    "                                if not msg_parts or (len(msg_parts) == 1 and not msg_parts[0].get('text', '').strip()):\n"
+    "                                    prior_output.pop()\n"
+    "                            output = []\n"
+    "                            await stream_body_handler(res, new_form_data)\n"
+    "                            output[:0] = prior_output\n"
+    "                            prior_output = []\n"
+    "                        else:\n"
+    "                            # TOOL_LOOP_ERRORS_UNIFIED: handle non-streaming error (ContextWindowExceeded, rate limit, etc)\n"
+    "                            _err_detail = None\n"
+    "                            try:\n"
+    "                                if hasattr(res, 'body') and isinstance(res.body, bytes):\n"
+    "                                    _resp_data = json.loads(res.body.decode('utf-8', 'replace'))\n"
+    "                                    if 'error' in _resp_data:\n"
+    "                                        _err_obj = _resp_data['error']\n"
+    "                                        _err_detail = _err_obj.get('message') or _err_obj.get('detail') or str(_err_obj) if isinstance(_err_obj, dict) else str(_err_obj)\n"
+    "                            except Exception:\n"
+    "                                _err_detail = f'Non-streaming error response: {type(res).__name__}'\n"
+    "                            if _err_detail:\n"
+    "                                log.error('NON_STREAM_ERROR: chat=%s iter=%d error=%s',\n"
+    "                                    metadata.get('chat_id', '')[:8], tool_call_retries, _err_detail)\n"
+    "                                if 'Model not found' in _err_detail:\n"
+    "                                    _err_detail = '" + _BUDGET_MSG + "'\n"
+    "                                # Keep only message items (text the user already saw)\n"
+    "                                _msg_items = [item for item in _saved_output if item.get('type') == 'message']\n"
+    "                                _msg_items.append({'type': 'message', 'id': '', 'status': 'completed', 'role': 'assistant', 'content': [{'type': 'output_text', 'text': f'\\n\\n---\\n**" + _ERROR_LABEL + ":** {_err_detail[:1000]}'}]})\n"
+    "                                output[:] = _msg_items\n"
+    "                                try:\n"
+    "                                    await event_emitter({'type': 'chat:message:error', 'data': {'error': {'content': _err_detail}}})\n"
+    "                                    await event_emitter({'type': 'chat:completion', 'data': {'content': serialize_output(output), 'output': output}})\n"
+    "                                except Exception:\n"
+    "                                    pass\n"
+    "                            break\n"
+    "                    except Exception as e:\n"
+    "                        # TOOL_LOOP_ERRORS_UNIFIED: restore clean output + show error; FIX_TOOL_LOOP_ERRORS\n"
+    "                        import traceback as _tb\n"
+    "                        _msg_items = [item for item in _saved_output if item.get('type') == 'message']\n"
+    "                        _err_mod = getattr(type(e), '__module__', '') or ''\n"
+    "                        _is_transport = 'aiohttp' in _err_mod or isinstance(e, (ConnectionError, TimeoutError, OSError))\n"
+    "                        if _is_transport:\n"
+    "                            log.warning('TRANSPORT_ERROR: chat=%s iter=%d error=%s',\n"
+    "                                metadata.get('chat_id', '')[:8], tool_call_retries, e)\n"
+    "                            _ui_err = '" + _TRANSPORT_MSG + "'\n"
+    "                        else:\n"
+    "                            log.error('TOOL_LOOP_ERROR: chat=%s iter=%d error=%s\\n%s',\n"
+    "                                metadata.get('chat_id', '')[:8], tool_call_retries, e, _tb.format_exc())\n"
+    "                            _ui_err = str(e)[:1000]\n"
+    "                            if 'Model not found' in _ui_err:\n"
+    "                                _ui_err = '" + _BUDGET_MSG + "'\n"
+    "                        try:\n"
+    "                            _msg_items.append({'type': 'message', 'id': '', 'status': 'completed', 'role': 'assistant', 'content': [{'type': 'output_text', 'text': f'\\n\\n---\\n**" + _ERROR_LABEL + ":** {_ui_err}'}]})\n"
+    "                            output[:] = _msg_items\n"
+    "                            await event_emitter({'type': 'chat:message:error', 'data': {'error': {'content': _ui_err}}})\n"
+    "                            await event_emitter({'type': 'chat:completion', 'data': {'content': serialize_output(output), 'output': output}})\n"
+    "                        except Exception:\n"
+    "                            pass\n"
+    "                        break\n"
+    "\n"
+    "                if DETECT_CODE_INTERPRETER:"
+)
+
+# ============================================================
+# Mod 2: Code interpreter catch -> log.error + UI error display
+# v0.9.1: `title = await Chats.get_chat_title_by_id(...)` — async-ified
+# v0.9.3+: title wrapped in multi-line parenthesized ternary (channel: guard)
+# ============================================================
+SEARCH_CODE_INTERP = """\
+                        except Exception as e:
+                            log.debug(e)
+                            break
+
+                # Mark all in-progress items as completed
+                for item in output:
+                    if item.get('status') == 'in_progress':
+                        item['status'] = 'completed'
+
+                title = (
+                    await Chats.get_chat_title_by_id(metadata['chat_id'])
+                    if not metadata['chat_id'].startswith('channel:')
+                    else ''
+                )"""
+
+REPLACE_CODE_INTERP = (
+    "                        except Exception as e:\n"
+    "                            import traceback as _tb  # TOOL_LOOP_ERRORS_UNIFIED; FIX_TOOL_LOOP_ERRORS\n"
+    "                            log.error('CODE_INTERP_ERROR: chat=%s iter=%d error=%s\\n%s',\n"
+    "                                metadata.get('chat_id', '')[:8], retries, e, _tb.format_exc())\n"
+    "                            try:\n"
+    "                                output.append({'type': 'message', 'id': '', 'status': 'completed', 'role': 'assistant', 'content': [{'type': 'output_text', 'text': f'\\n\\n---\\n**" + _ERROR_LABEL + ":** {str(e)[:1000]}'}]})\n"
+    "                                await event_emitter({'type': 'chat:completion', 'data': {'content': serialize_output(output), 'output': output}})\n"
+    "                            except Exception:\n"
+    "                                pass\n"
+    "                            break\n"
+    "\n"
+    "                # Mark all in-progress items as completed\n"
+    "                for item in output:\n"
+    "                    if item.get('status') == 'in_progress':\n"
+    "                        item['status'] = 'completed'\n"
+    "\n"
+    "                title = (\n"
+    "                    await Chats.get_chat_title_by_id(metadata['chat_id'])\n"
+    "                    if not metadata['chat_id'].startswith('channel:')\n"
+    "                    else ''\n"
+    "                )"
+)
+
+# ============================================================
+# Mod 3: SSE parse catch -> log.error with context
+# ============================================================
+SEARCH_SSE = """\
+                        except Exception as e:
+                            done = 'data: [DONE]' in line
+                            if done:
+                                pass
+                            else:
+                                log.debug(f'Error: {e}')
+                                continue"""
+
+REPLACE_SSE = """\
+                        except Exception as e:
+                            done = 'data: [DONE]' in line
+                            if done:
+                                pass
+                            else:
+                                log.error('SSE_PARSE_ERROR: chat=%s line=%.200s error=%s',  # TOOL_LOOP_ERRORS_UNIFIED; FIX_TOOL_LOOP_ERRORS
+                                    metadata.get('chat_id', '')[:8], str(line)[:200], e)
+                                continue"""
+
+# ============================================================
+# Mod 4: Done emit try/except + background_tasks_handler wrapper
+# v0.9.1: two new lines inserted between `await background_tasks_handler(ctx)`
+# and `except asyncio.CancelledError:` — `ctx['assistant_message'] = {...}` +
+# `await outlet_filter_handler(ctx)`. Both must be preserved inside the wrap.
+# ============================================================
+SEARCH_DONE_BG = """\
+                await event_emitter(
+                    {
+                        'type': 'chat:completion',
+                        'data': data,
+                    }
+                )
+
+                await background_tasks_handler(ctx)
+                ctx['assistant_message'] = {
+                    'content': serialize_output(output),
+                    'output': output,
+                    **({'usage': usage} if usage else {}),
+                }
+                await outlet_filter_handler(ctx)
+            except asyncio.CancelledError:"""
+
+REPLACE_DONE_BG = """\
+                try:  # TOOL_LOOP_ERRORS_UNIFIED: wrap done emit; FIX_TOOL_LOOP_ERRORS
+                    await event_emitter(
+                        {
+                            'type': 'chat:completion',
+                            'data': data,
+                        }
+                    )
+                except Exception as _done_err:
+                    log.error('DONE_EMIT_ERROR: chat=%s error=%s',
+                        metadata.get('chat_id', '')[:8], _done_err)
+
+                try:
+                    await background_tasks_handler(ctx)
+                    ctx['assistant_message'] = {
+                        'content': serialize_output(output),
+                        'output': output,
+                        **({'usage': usage} if usage else {}),
+                    }
+                    await outlet_filter_handler(ctx)
+                except Exception as _bg_err:
+                    log.error('BACKGROUND_TASK_ERROR: chat=%s error=%s',  # TOOL_LOOP_ERRORS_UNIFIED
+                        metadata.get('chat_id', '')[:8], _bg_err)
+            except asyncio.CancelledError:"""
+
+# ============================================================
+# Mod 5: TOOL_LOOP_ITER lifecycle logging
+# ============================================================
+SEARCH_ITER = """\
+                while len(tool_calls) > 0 and tool_call_retries < CHAT_RESPONSE_MAX_TOOL_CALL_RETRIES:
+                    tool_call_retries += 1"""
+
+REPLACE_ITER = """\
+                while len(tool_calls) > 0 and tool_call_retries < CHAT_RESPONSE_MAX_TOOL_CALL_RETRIES:
+                    tool_call_retries += 1
+                    log.debug('TOOL_LOOP_ITER: chat=%s iter=%d pending_tc=%d',  # TOOL_LOOP_ERRORS_UNIFIED; FIX_TOOL_LOOP_ERRORS
+                        metadata.get('chat_id', '')[:8], tool_call_retries, len(tool_calls))"""
+
+
+def apply_patch():
+    if not os.path.exists(MIDDLEWARE_PATH):
+        print(
+            f"ERROR: fix_tool_loop_errors target file {MIDDLEWARE_PATH} not found. "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+
+    with open(MIDDLEWARE_PATH, "r", encoding="utf-8") as f:
+        content = f.read()
+
+    if PATCH_MARKER in content or NEW_PATCH_MARKER in content:
+        print(f"ALREADY PATCHED: {MIDDLEWARE_PATH} contains {PATCH_MARKER}")
+        return True
+
+    # v0.9.1 -> v0.9.2 backward-compat shim: v0.9.2 upstream inserted a new
+    # `'metadata': metadata,` key into the first `new_form_data = {` block. The
+    # SEARCH_TOOL_LOOP anchor targets the v0.9.2 shape; for v0.9.1 input we
+    # inject the missing key in-memory so the single SEARCH matches both
+    # upstream versions. No-op on v0.9.2 (V091_SHIM does not match there).
+    V091_SHIM = "                            'stream': True,\n                        }\n\n                        if ENABLE_RESPONSES_API_STATEFUL"
+    V092_SHIM = "                            'stream': True,\n                            'metadata': metadata,\n                        }\n\n                        if ENABLE_RESPONSES_API_STATEFUL"
+    if V091_SHIM in content and V092_SHIM not in content:
+        content = content.replace(V091_SHIM, V092_SHIM, 1)
+
+    # Mod 1/5: tool_loop
+    if SEARCH_TOOL_LOOP not in content:
+        print(
+            f"ERROR: fix_tool_loop_errors anchor 1/5 (tool_loop) not found in {MIDDLEWARE_PATH} "
+            "— upstream may have refactored the tool-retry try/except. "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+    content = content.replace(SEARCH_TOOL_LOOP, REPLACE_TOOL_LOOP, 1)
+    print("  [1/5] Tool loop: save/restore + transport + non-stream + Model not found + chat:message:error")
+
+    # Mod 2/5: code_interp (v0.9.1: await-ified Chats.get_chat_title_by_id)
+    if SEARCH_CODE_INTERP not in content:
+        print(
+            f"ERROR: fix_tool_loop_errors anchor 2/5 (code_interp) not found in {MIDDLEWARE_PATH} "
+            "— upstream may have refactored the code-interpreter except block or the "
+            "await Chats.get_chat_title_by_id call. "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+    content = content.replace(SEARCH_CODE_INTERP, REPLACE_CODE_INTERP, 1)
+    print("  [2/5] Code interpreter: log.error + UI error display")
+
+    # Mod 3/5: sse
+    if SEARCH_SSE not in content:
+        print(
+            f"ERROR: fix_tool_loop_errors anchor 3/5 (sse) not found in {MIDDLEWARE_PATH} "
+            "— upstream may have refactored the SSE parse except block. "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+    content = content.replace(SEARCH_SSE, REPLACE_SSE, 1)
+    print("  [3/5] SSE parse: log.error with context")
+
+    # Mod 4/5: done_bg
+    if SEARCH_DONE_BG not in content:
+        print(
+            f"ERROR: fix_tool_loop_errors anchor 4/5 (done_bg) not found in {MIDDLEWARE_PATH} "
+            "— upstream may have refactored the background_tasks_handler block "
+            "(v0.9.1 inserted assistant_message + outlet_filter_handler). "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+    content = content.replace(SEARCH_DONE_BG, REPLACE_DONE_BG, 1)
+    print("  [4/5] Done emit wrapped + background_tasks_handler wrapped")
+
+    # Mod 5/5: iter
+    if SEARCH_ITER not in content:
+        print(
+            f"ERROR: fix_tool_loop_errors anchor 5/5 (iter) not found in {MIDDLEWARE_PATH} "
+            "— upstream may have refactored the tool-retry while-loop start. "
+            "Refusing to produce a silently-broken image.",
+            file=sys.stderr,
+        )
+        sys.exit(1)
+    content = content.replace(SEARCH_ITER, REPLACE_ITER, 1)
+    print("  [5/5] TOOL_LOOP_ITER lifecycle logging")
+
+    with open(MIDDLEWARE_PATH, "w", encoding="utf-8") as f:
+        f.write(content)
+
+    print("PATCHED: fix_tool_loop_errors applied successfully.")
+    print("  Log markers: TOOL_LOOP_ERROR, TRANSPORT_ERROR, NON_STREAM_ERROR,")
+    print("               CODE_INTERP_ERROR, SSE_PARSE_ERROR, DONE_EMIT_ERROR,")
+    print("               BACKGROUND_TASK_ERROR, TOOL_LOOP_ITER")
+    return True
+
+
+if __name__ == "__main__":
+    print("Applying unified tool loop errors patch to middleware.py...")
+    success = apply_patch()
+    sys.exit(0 if success else 1)

package/package.json

@@ -0,0 +1,39 @@
+{
+  "name": "@mseep/open-computer-use",
+  "mcpName": "io.github.Wide-Moat/open-computer-use",
+  "version": "1.0.0",
+  "description": "Open Computer Use - Node.js Dependencies",
+  "license": "FSL-1.1-Apache-2.0",
+  "engines": {
+    "node": ">=22.20.0",
+    "npm": ">=10.9.3"
+  },
+  "dependencies": {
+    "@mermaid-js/mermaid-cli": "^11.12.0",
+    "docx": "^9.5.1",
+    "graphviz": "^0.0.9",
+    "markdown-pdf": "^11.0.0",
+    "markdown-toc": "^1.2.0",
+    "markdownlint-cli": "^0.45.0",
+    "markdownlint-cli2": "^0.18.1",
+    "marked": "^16.4.0",
+    "pdf-lib": "^1.17.1",
+    "pdfjs-dist": "^5.4.296",
+    "playwright": "1.57.0",
+    "pptxgenjs": "^4.0.1",
+    "react": "^19.2.0",
+    "react-dom": "^19.2.0",
+    "react-icons": "^5.5.0",
+    "remark-cli": "^12.0.1",
+    "remark-preset-lint-recommended": "^7.0.1",
+    "sharp": "^0.34.4",
+    "ts-node": "^10.9.2",
+    "tsx": "^4.20.6",
+    "typescript": "^5.9.3"
+  },
+  "keywords": [
+    "mseep",
+    "mcp-server"
+  ],
+  "publisher": "mseep"
+}

package/requirements.txt

@@ -0,0 +1,112 @@
+# AI Computer Use - Python Dependencies
+# Python 3.12.3
+
+# Core document processing
+python-docx==1.2.0
+python-pptx==1.0.2
+openpyxl==3.1.5
+xlsxwriter==3.2.9
+xlrd==2.0.1
+
+# PDF processing
+pypdf==5.9.0
+pdfplumber==0.11.9
+pypdfium2==4.30.0
+PyMuPDF==1.24.10
+pikepdf==9.11.0
+pdfkit==1.0.0
+pdfminer.six==20251230
+pdf2image==1.17.0
+reportlab==4.4.10
+
+# Image processing
+pillow==12.2.0
+opencv-python==4.11.0.86
+opencv-contrib-python==4.11.0.86
+opencv-python-headless==4.11.0.86
+imageio==2.37.0
+imageio-ffmpeg==0.6.0
+scikit-image==0.25.2
+Wand==0.6.13
+
+# OCR and text extraction
+pytesseract==0.3.13
+tabula-py==2.10.0
+camelot-py==1.0.9
+
+# Web and HTTP
+requests==2.32.5
+urllib3==2.6.3
+httplib2==0.20.4
+certifi==2025.10.5
+beautifulsoup4==4.14.3
+soupsieve==2.8
+
+# LLM SDKs
+openai>=2.20.0
+
+# Markdown and documentation
+markdown-it-py==4.0.0
+markdownify==1.2.0
+markitdown==0.1.3
+marko==2.2.1
+mistune==3.1.4
+grip==4.6.2
+
+# Data processing
+pandas==2.3.3
+numpy==2.3.3
+scipy==1.16.2
+scikit-learn==1.7.2
+
+# Plotting and visualization
+matplotlib==3.10.7
+seaborn==0.13.2
+
+# XML processing
+lxml==6.0.2
+et_xmlfile==2.0.0
+defusedxml==0.7.1
+odfpy==1.4.1
+
+# Machine Learning
+jax==0.7.2
+jaxlib==0.7.2
+# tensorflow-lite==2.13.0 - does NOT exist on PyPI, removed
+onnxruntime==1.23.1
+ml_dtypes==0.5.3
+mediapipe==0.10.14
+magika==0.6.2
+
+# Web automation
+playwright==1.57.0
+pyee==13.0.0
+
+# Office/LibreOffice integration
+unoserver==3.4
+pyoo==1.4
+
+# CLI and utilities
+click==8.3.0
+colorama==0.4.6
+coloredlogs==15.0.1
+humanfriendly==10.0
+tabulate==0.9.0
+psutil==7.2.2
+
+# Dev tools
+Flask==3.1.3
+Werkzeug==3.1.3
+
+# Other dependencies
+cryptography==46.0.6
+PyJWT==2.12.1
+python-dateutil==2.9.0.post0
+pytz==2025.2
+tzdata==2025.2
+python-dotenv==1.1.1
+PyYAML==6.0.3
+pyyaml_env_tag==1.1
+wrapt==1.17.3
+Deprecated==1.2.18
+six==1.17.0