feat(fase2): feed debate converged output into candidate generation

src/portfolio_optimiser/run.py

@@ -27,6 +27,7 @@ from __future__ import annotations
 
 from collections.abc import Callable
 from dataclasses import dataclass
+from typing import Any
 
 from agent_framework import BaseChatClient, SessionContext
 
@@ -56,13 +57,35 @@ from portfolio_optimiser.workflow import fresh_workflow
 @dataclass(frozen=True)
 class RunResult:
     """The outcome of one project run: the validated/rejected proposal, its first-class
-    provenance, the captured (Layer-2) verdict, the ExpeL hits surfaced for it, and the store."""
+    provenance, the captured (Layer-2) verdict, the ExpeL hits surfaced for it, the store, and
+    the debate's converged output that the candidate was generated from (F1 traceability)."""
 
     outcome: ValidatedProposal | Rejection
     provenance: ProvenanceStamp
     verdict: Verdict
     retrieved: list[Verdict]
     store: VerdictStore
+    debate_output: str
+
+
+# The GroupChat orchestrator emits this terminal notice as a participant output; it is NOT the
+# debate's substantive result, so the F1 extractor filters it out (verified against MAF 1.9.0).
+_ORCH_NOTICE = "reached the maximum number of rounds"
+
+
+def _debate_text(result: Any) -> str:
+    """Extract the proposer's converged output from a ``WorkflowRunResult`` (surfaced by
+    ``output_from=[proposer]``), excluding the orchestrator's termination notice. Returns ``""``
+    when no substantive participant text is present."""
+    texts: list[str] = []
+    for out in result.get_outputs():
+        text = out if isinstance(out, str) else getattr(out, "text", None)
+        if not text:
+            continue
+        if getattr(out, "author_name", None) == "group_chat_orchestrator" or _ORCH_NOTICE in text:
+            continue
+        texts.append(text)
+    return texts[-1] if texts else ""
 
 
 def _project_by_id(project_id: str) -> Project:
@@ -136,10 +159,14 @@ async def run_project(
         tools=[retrieval_tool],
         middleware=[budget_mw],
     )
-    await debate.run(f"Find a cost-saving measure for {project.id}.\nContext:\n{context}")
+    result = await debate.run(f"Find a cost-saving measure for {project.id}.\nContext:\n{context}")
+    # F1: the candidate must derive from the DEBATE. Feed the proposer's converged output into
+    # generation (retrieval context is the last-resort fallback only).
+    debate_output = _debate_text(result)
+    gen_context = debate_output or context
 
     # 5. Structured candidate -> blocking validation; token bound = the meter in this loop.
-    outcome = await generate_via_llm(factory("proposer"), project, context, meter)
+    outcome = await generate_via_llm(factory("proposer"), project, gen_context, meter)
     proposal = outcome.proposal
 
     # 6. First-class provenance stamp (authoritative; independent of MAF Annotation).
@@ -168,7 +195,12 @@ async def run_project(
         notify(verdict)
 
     return RunResult(
-        outcome=outcome, provenance=stamp, verdict=verdict, retrieved=retrieved, store=store
+        outcome=outcome,
+        provenance=stamp,
+        verdict=verdict,
+        retrieved=retrieved,
+        store=store,
+        debate_output=debate_output,
     )
 
 

tests/test_vertical_slice_e2e.py

@@ -138,6 +138,37 @@ async def test_wiring_budget_middleware_and_retrieval_tool(
     assert any(isinstance(t, FunctionTool) for t in tools)
 
 
+async def test_g_validated_proposal_derives_from_debate(
+    docs_dir, make_client_factory, fresh_store, monkeypatch
+) -> None:
+    """F1: the candidate fed to generation derives from the DEBATE, not just retrieval. Spy
+    the context passed to generate_via_llm; the proposer's converged output ('Reduce scope',
+    present in _VALID but NOT in the docs_dir fixture) must reach generation. Deleting the
+    debate->generation wiring makes this fail."""
+    from portfolio_optimiser import run as run_mod
+
+    captured: dict[str, str] = {}
+    real_generate = run_mod.generate_via_llm
+
+    async def spy_generate(chat_client, project, context, meter, **kw):
+        captured["context"] = context
+        return await real_generate(chat_client, project, context, meter, **kw)
+
+    monkeypatch.setattr(run_mod, "generate_via_llm", spy_generate)
+    result = await run_project(
+        "FV42-GSV-E1",
+        "local",
+        docs_dir=docs_dir,
+        verdict_input=_VI,
+        client_factory=make_client_factory(_VALID),
+        store=fresh_store,
+    )
+    debate_marker = "Reduce scope"  # in _VALID (proposer output); NOT in docs_dir
+    assert result.debate_output and debate_marker in result.debate_output
+    assert debate_marker in captured["context"]  # debate output reached generation
+    assert isinstance(result.outcome, ValidatedProposal)
+
+
 async def test_f_malformed_contract_raises_before_any_chat(docs_dir, make_client_factory) -> None:
     calls = {"n": 0}
     base = make_client_factory(_VALID)