vodorod/Dorod-Sky
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Do not enforce regenerating tasks when url is passed (#4144)

Browse Source
This commit is contained in:
Shuchang Zheng
2025-11-29 13:49:26 -08:00
committed by GitHub
parent 76735cd8a6
commit cb5ee92f59
2 changed files with 14 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
skyvern/forge/sdk/routes/agent_protocol.py
View File

@@ -164,17 +164,19 @@ async def run_task(
data_extraction_schema = run_request.data_extraction_schema
navigation_goal = run_request.prompt
navigation_payload = None
task_generation = await task_v1_service.generate_task(
user_prompt=run_request.prompt,
organization=current_org,
)
url = url or task_generation.url
navigation_goal = task_generation.navigation_goal or run_request.prompt
if run_request.engine in CUA_ENGINES:
navigation_goal = run_request.prompt
navigation_payload = task_generation.navigation_payload
data_extraction_goal = task_generation.data_extraction_goal
data_extraction_schema = data_extraction_schema or task_generation.extracted_information_schema
if not url:
task_generation = await task_v1_service.generate_task(
user_prompt=run_request.prompt,
organization=current_org,
)
# What if it's a SDK request with browser_session_id?
url = task_generation.url
navigation_goal = task_generation.navigation_goal or run_request.prompt
if run_request.engine in CUA_ENGINES:
navigation_goal = run_request.prompt
navigation_payload = task_generation.navigation_payload
data_extraction_goal = task_generation.data_extraction_goal
data_extraction_schema = data_extraction_schema or task_generation.extracted_information_schema
task_v1_request = TaskRequest(
title=run_request.title,

2
skyvern/services/task_v1_service.py
View File

@@ -47,7 +47,7 @@ async def generate_task(user_prompt: str, organization: Organization) -> TaskGen
llm_prompt = prompt_engine.load_prompt("generate-task", user_prompt=user_prompt)
try:
llm_response = await app.LLM_API_HANDLER(
llm_response = await app.SECONDARY_LLM_API_HANDLER(
prompt=llm_prompt, prompt_name="generate-task", organization_id=organization.organization_id
)
parsed_task_generation_obj = TaskGenerationBase.model_validate(llm_response)