Commit Graph

339 Commits

Author SHA1 Message Date
Dillon DuPont
50d8521e3e added computer.start(), computer.restart() 2025-10-09 13:27:24 -04:00
Dillon DuPont
9595f5affa updated api examples 2025-10-09 13:12:55 -04:00
Dillon DuPont
b29f89597f added shared provider vm type 2025-10-09 12:36:43 -04:00
Dillon DuPont
0ede822990 Add /start, /stop, and /restart to cloud provider 2025-10-09 12:01:45 -04:00
Dillon DuPont
d0fefd3cd7 add google-genai to pyproject.toml 2025-10-09 11:59:22 -04:00
Dillon DuPont
ef28a64b8f Add gemini CUA loop 2025-10-07 17:23:33 -04:00
Dillon DuPont
78901aed48 Add --provider and --prompt-file to the CLI 2025-10-07 11:04:21 -04:00
bowman
1fe41d57f4 update hud in agent pyproject.toml 2025-10-06 20:24:21 -07:00
Adam
3274cfafe7 add concurrent session management and resource isolation
Implement concurrent session management for MCP server with:

- SessionManager with computer instance pooling
- Concurrent task execution support
- New tools: get_session_stats, cleanup_session
- Graceful shutdown and resource cleanup
- Fix nested asyncio event loop issues
- Add comprehensive tests and documentation

Enables multiple concurrent clients with proper resource isolation
while maintaining backward compatibility.
2025-10-06 18:37:10 -04:00
ddupont
9f18f9eeaa Merge pull request #437 from trycua/feat/moondream3
Add moondream3 model for CUA
2025-10-06 13:07:33 -04:00
ddupont
9102cd1490 Merge pull request #439 from trycua/feat/windows-cloud
Support Windows cloud VMs in Computer SDK
2025-10-03 12:15:41 -04:00
Dillon DuPont
1cacd281ad fixed python RPC on windows 2025-10-03 12:12:40 -04:00
Dillon DuPont
4dfc8de46b Disable watchdog on windows cloud instances 2025-10-03 10:09:17 -04:00
Dillon DuPont
1e94b5d8b4 Added working moondream3 agent 2025-10-02 11:07:11 -04:00
Dillon DuPont
0b3c677205 added moondream3 agent loop 2025-10-02 10:57:06 -04:00
Adam
671845001c update README with corrected MCP config and clarified usage
- Updated the README to clarify how to configure cua-mcp-server for Claude Desktop and other MCP clients.
- Fixed the example development configuration to use an absolute/expanded path rather than `~`, preventing issues with Claude’s sandboxed execution.
- Aligned instructions with the current start_mcp_server.sh behavior, ensuring the correct virtual environment and PYTHONPATH are always set.
- Improved wording for consistency across installation, development, and usage sections.
2025-10-01 11:49:17 -04:00
Adam
37864faef2 Fix MCP server startup and tool schema issues in server.py and start_mcp_server.sh
Changes
- Refactored `server.py`:
  - Delayed server instantiation to avoid double-import issues and RuntimeWarning.
  - Reworked `run_cua_task` and `run_multi_cua_tasks` to yield structured output
    (text + screenshots) in a Pydantic-compatible format.
  - Added helper functions for normalising message content, extracting text,
    and serialising tool outputs.
  - Improved logging and error handling for clearer debugging.

- Updated `start_mcp_server.sh`:
  - Ensures Python path points to repo-local `.venv`.
  - Sets `PYTHONPATH` correctly for all `libs/python/*` packages.
  - Added clear debug output for python binary, repo dir, and PYTHONPATH.

Why
- Fixed `PydanticSchemaGenerationError` when returning `Image` objects from
  tool functions.
- Prevented `FastMCP.__init__()` errors by aligning with supported arguments.
- Removed duplicate module loading at startup that triggered RuntimeWarnings.
- Improved clarity of server startup logs for easier integration with Claude Desktop.

Outcome
The MCP server now starts cleanly, registers all tools, and stays connected
with Claude Desktop without schema errors or double-import issues.
2025-09-30 18:52:04 -04:00
Dillon DuPont
c892f4ecea Change anthropic predict_click logic 2025-09-25 15:53:20 -07:00
Adam
65263112cd Stream MCP responses instead of buffering 2025-09-25 16:07:53 -04:00
Dillon DuPont
f4be6538ff Hide setup window 2025-09-22 09:29:09 -04:00
Dillon DuPont
70f9f968cb Fixed broken winsandbox_example, added venv to provider setup 2025-09-22 08:59:07 -04:00
Dillon DuPont
c51692cd0c Fix image must have : error 2025-09-22 08:40:18 -04:00
Dillon DuPont
28f097f987 Merge branch 'main' of https://github.com/trycua/cua 2025-09-19 19:23:16 -04:00
Dillon DuPont
1346feb125 Add postponed annotations to internvl.py 2025-09-19 19:23:13 -04:00
ddupont
3ac566f53a Merge pull request #416 from trycua/feat/auth-ttl
[Agent] Add CUA_AUTH_TTL_SECONDS env, change default from 5s to 60s
2025-09-19 19:22:03 -04:00
Dillon DuPont
26ff9e1105 Add CUA_AUTH_TTL_SECONDS env, change default from 5s to 60s 2025-09-18 12:40:59 -04:00
Dillon DuPont
821bd03e48 remove extra prints 2025-09-18 11:27:50 -04:00
Dillon DuPont
03d7806549 Fixed invalid trajectory names on ollama 2025-09-18 10:52:45 -04:00
Dillon DuPont
6ddddf8f88 fix internVL inference 2025-09-16 12:56:07 -04:00
Dillon DuPont
9147e8eeaf Added "cua-agent[internvl-hf]" dep 2025-09-16 12:02:07 -04:00
Dillon DuPont
c5bbd4611a add qwen2_5_vl.py 2025-09-15 16:29:26 -04:00
Dillon DuPont
7a7de5d50f add holo models 2025-09-15 16:10:54 -04:00
Dillon DuPont
ca564b2436 Merge branch 'main' into models/opencua 2025-09-15 15:11:15 -04:00
James Murdza
b4b45e5b8b Upgrade HUD SDK to 0.4.26 2025-09-14 00:34:51 -04:00
James Murdza
8096fbfd34 Upgrade HUD SDK to 0.4.25 2025-09-13 22:43:12 -04:00
Dillon DuPont
c58ff55969 Added agent tool filtering 2025-09-12 20:12:29 -04:00
Dillon DuPont
fc0f10aaf9 Ignore extra computers when running evals 2025-09-12 14:50:18 -04:00
Dillon DuPont
ae22572658 Merge branch 'main' of https://github.com/trycua/cua 2025-09-12 13:58:22 -04:00
Dillon DuPont
fad293957d Pin HUD version 2025-09-12 13:57:28 -04:00
Dillon DuPont
cf95646503 Merge branch 'main' into models/opencua 2025-09-12 13:30:21 -04:00
Dillon DuPont
58807378dd Added internVL 2025-09-12 13:30:09 -04:00
Dillon DuPont
faf531825e Fixed error during response call 2025-09-12 12:32:03 -04:00
Dillon DuPont
b3040306b8 Fixing bugs 2025-09-12 12:06:36 -04:00
Dillon DuPont
b69943121d Fixed KeyError 2025-09-12 11:29:40 -04:00
Dillon DuPont
f795660f75 Upgraded HUD impl. to support custom tools 2025-09-12 11:14:03 -04:00
Dillon DuPont
2f28c3a2ce Added missing file 2025-09-10 14:57:44 -04:00
Dillon DuPont
17d6709629 added simple guide for customizing computeragent 2025-09-09 10:55:57 -04:00
ddupont
da9af2e0fd Merge pull request #390 from onel/reference-docs-20250901_145129
Reference documentation batch
2025-09-05 11:17:35 -04:00
Dillon DuPont
2dfaf0047d Fix multimodal user inputs in the anthropic loop 2025-09-04 16:15:33 -04:00
Dillon DuPont
8c49ce317a Fixed X/Y scrolling on windows 2025-09-04 16:09:47 -04:00