MCPcopy Create free account
hub / github.com/ScaleML/AgentSPEX / cli

Function cli

src/harness/ephemeral/cli.py:100–122  ·  view source on GitHub ↗

Run the agentic loop against a running sandbox VM. Args: model: LLM model to use. Defaults to claude-sonnet-4-5. mcp_url: MCP server URL. Defaults to http://localhost:7002/mcp. output_dir: Output directory. Defaults to ``workspace_persistent/outputs/agentic_l

(
    model="claude-sonnet-4-5",
    mcp_url="http://localhost:7002/mcp",
    output_dir=None,
    temperature=0.2,
)

Source from the content-addressed store, hash-verified

98
99
100def cli(
101 model="claude-sonnet-4-5",
102 mcp_url="http://localhost:7002/mcp",
103 output_dir=None,
104 temperature=0.2,
105):
106 """Run the agentic loop against a running sandbox VM.
107
108 Args:
109 model: LLM model to use. Defaults to claude-sonnet-4-5.
110 mcp_url: MCP server URL. Defaults to http://localhost:7002/mcp.
111 output_dir: Output directory. Defaults to
112 ``workspace_persistent/outputs/agentic_loop`` under the project root.
113 temperature: LLM sampling temperature. Defaults to 0.2.
114 """
115 from harness.agentic_loop.run_loop import run_loop
116
117 run_loop(
118 model=model,
119 mcp_url=mcp_url,
120 output_dir=output_dir,
121 temperature=temperature,
122 )
123
124
125def main():

Callers

nothing calls this directly

Calls 1

run_loopFunction · 0.90

Tested by

no test coverage detected