Agent-S

Agent-S copied to clipboard

Fixed this: ``` Traceback (most recent call last): File "/config/Desktop/Agent-S/./gui_agents/s2/cli_app.py", line 276, in main() File "/config/Desktop/Agent-S/./gui_agents/s2/cli_app.py", line 222, in main "engine_type": args.model_provider, ^^^^^^^^^^^^^^^^^^^ AttributeError: 'Namespace' object has no attribute 'model_provider'

DaWe35

does gemini support embedding? Must I use openai embedding?

1

does gemini support embedding? Must I use openai embedding? ^^^^^^^^^^^^^^ File "d:\github\pyprojects\trae\ai agents\agent-s\Agent-S\gui_agents\s2\core\knowledge.py", line 34, in __init__ self.embedding_engine = OpenAIEmbeddingEngine() ^^^^^^^^^^^^^^^^^^^^^^^ File "d:\github\pyprojects\trae\ai agents\agent-s\Agent-S\gui_agents\s2\core\engine.py", line 32, in __init__ raise ValueError(...

diabloaltman

Support parsing DAG JSON from <json>...</json> tags and json code blocks to improve the robustness of the parse_dag function.

1

During testing with the Qwen2.5-VL model, the JSON output is mostly wrapped inside json code blocks rather than the expected ... tag format. Currently, the parse_dag function only extracts JSON...

iejzh

Gemini models as grounding agent x,y point outputs are inverted and normalized.

3

Using the framework with Gemini 2.5 Flash as the base and grounding agents, I found it difficult to get the agent to correctly set the location of various icons or...

Tsukalos

**Context and Purpose:** This PR automatically remediates a security vulnerability: - **Description:** Detected the use of eval(). eval() can be dangerous if used to evaluate dynamic content. If this content...

kira-offgrid

**Context and Purpose:** This PR automatically remediates a security vulnerability: - **Description:** Detected the use of eval(). eval() can be dangerous if used to evaluate dynamic content. If this content...

kira-offgrid

Reproduction details of Agent S3 results in Paper

15

Thanks for the great work! We are very interested in reproducing the results from the paper. By default, we are using the GPT-5 Azure API and vLLM hosted with Uitars...

linyq17

Task execution time (latency) of Agent-S

1

Hello ! for some quick prototyping I would like to have AI Agent that works just fast on my PC. By fast, I mean, let say, max 15 seconds for...

kubaaa2111

Any video usage demo ?

3

Hi, i had a hard time to setup the apps myself due to api issues. Just wondering if there any usage model could be linked to readme showing how to...

lucasjinreal

Failed to buffer the request body: length limit exceeded.

9

I am using UI-TARS-1.5-7B hosted on **Hugging Face Inference Endpoints**, running the command **in Ubuntu**: ``` agent_s \ --provider openai \ --model gpt-5-2025-08-07 \ --ground_provider huggingface \ --ground_url [MY_URL] \...

FFchopon