Class: Collavre::AiAgentService

Inherits:
Object
  • Object
show all
Defined in:
app/services/collavre/ai_agent_service.rb

Overview

Orchestrates AI agent execution for a task. Delegates specific concerns to specialized service objects:

  • AgentLifecycleManager: agent status, cancellation, heartbeats

  • ResponseStreamer: streaming content updates

  • ResponseFinalizer: comment finalization, review workflow

  • A2aDispatcher: agent-to-agent event dispatch

Constant Summary collapse

CANCEL_CHECK_INTERVAL =

Compatibility alias for constants moved to AgentLifecycleManager

AiAgent::AgentLifecycleManager::CANCEL_CHECK_INTERVAL

Instance Method Summary collapse

Constructor Details

#initialize(task) ⇒ AiAgentService

Returns a new instance of AiAgentService.



12
13
14
15
16
 
# File 'app/services/collavre/ai_agent_service.rb', line 12

def initialize(task)
  @task = task
  @agent = task.agent
  @context = task.trigger_event_payload
end

Instance Method Details

#call ⇒ Object 



18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
 
# File 'app/services/collavre/ai_agent_service.rb', line 18

def call
  Current.set(user: @agent) do
    log_action("start", { message: "Starting agent execution" })

    # Build context and messages
    @original_comment = find_original_comment
    messages_data = build_messages
    log_action("prompt_generated", { messages: messages_data[:messages] })

    # Prepare rendering context and prompts
    @creative = find_creative
    rendering_context = prepare_rendering_context
    system_prompt = render_system_prompt(rendering_context)

    # Resolve session context (decides what to send)
    resolved = resolve_session_context(messages_data, system_prompt)

    # Create placeholder comment if needed
    @reply_comment = create_reply_comment_if_needed

    # Initialize lifecycle manager
    @lifecycle_manager = AiAgent::AgentLifecycleManager.new(
      task: @task,
      agent: @agent,
      creative: @creative
    )

    # Initialize response streamer
    @streamer = AiAgent::ResponseStreamer.new(
      reply_comment: @reply_comment,
      creative: @creative
    )

    @lifecycle_manager.broadcast_status("thinking")

    # Execute AI chat with streaming
    @client = build_ai_client(resolved[:system_prompt])
    stream_response(@client, resolved)

    log_action("completion", { response: @streamer.content })

    # Finalize and dispatch (skip A2A for review-flow updates)
    finalized_comment = finalize_response
    dispatch_a2a(finalized_comment) unless @finalizer&.review_flow

    @lifecycle_manager.broadcast_status("idle")

    @streamer.content
  end
rescue ApprovalPendingError => e
  summary = generate_approval_summary(e)
  AiAgent::ApprovalHandler.new(
    task: @task, agent: @agent, context: @context,
    creative: @creative, reply_comment: @reply_comment
  ).handle(e, summary: summary)
  raise
rescue CancelledError
  handle_cancelled
  abort_openclaw_if_needed
  raise
end