fix(cua): increase default max_steps_per_phase to 25 — connect phase needs more room for first-launch UI navigation
This commit is contained in:
@@ -741,7 +741,7 @@ def run_onboarding_showcase(
|
||||
model: str = "gpt-5.4",
|
||||
include_ui_xml: bool = False,
|
||||
verbose: bool = True,
|
||||
max_steps_per_phase: int = 20,
|
||||
max_steps_per_phase: int = 25,
|
||||
) -> dict:
|
||||
"""Execute the full first-run onboarding journey.
|
||||
|
||||
@@ -1772,7 +1772,7 @@ Examples:
|
||||
|
||||
# Model / verbosity
|
||||
parser.add_argument("--model", default="gpt-4o", help="Vision model deployment name.")
|
||||
parser.add_argument("--max-steps", type=int, default=20, help="Max LLM steps per phase (showcase) or total (legacy).")
|
||||
parser.add_argument("--max-steps", type=int, default=25, help="Max LLM steps per phase (showcase) or total (legacy).")
|
||||
parser.add_argument("--include-xml", action="store_true", help="Include UI hierarchy XML in LLM context (more accurate, more tokens).")
|
||||
parser.add_argument("--quiet", action="store_true")
|
||||
|
||||
|
||||
Reference in New Issue
Block a user