diff --git a/roboco/services/gateway/choreographer.py b/roboco/services/gateway/choreographer.py index fd70bdb4..7978e9dc 100644 --- a/roboco/services/gateway/choreographer.py +++ b/roboco/services/gateway/choreographer.py @@ -322,10 +322,17 @@ class Choreographer: ) async def i_am_done(self, agent_id: UUID, task_id: UUID, notes: str) -> Envelope: - """Submit work for QA. Runs verify/push/PR/submit-qa sequentially as needed. + """Submit work for QA — strict path. - Each step gated by tracing/state preconditions; returns precise - remediation hints when prerequisites are missing. + Pre-gateway, the route layer enforced four field-level gates + (NOT_SELF_VERIFIED, NO_COMMITS, NO_PR, NO_PROGRESS) before + transitioning verifying → awaiting_qa. The strict gateway path + re-enforces those exactly: dev MUST have already committed, + pushed, opened a PR, reported progress, and self-verified before + i_am_done can submit. + + For the smart-catch-up convenience that auto-runs the chain + on the dev's behalf, see ``i_am_done_with_catchup``. """ t = await self.task.get(task_id) if t is None: @@ -337,7 +344,56 @@ class Choreographer: context_briefing=await self._briefing_for(agent_id, task_id), ) - # 1. Tracing-gate preconditions + # 1. Tracing-gate preconditions (progress / reflect / acceptance) + if rejection := await self._check_tracing_gates(agent_id, task_id, t): + return rejection + + # 2. Field-level gates (Gate Set E) — strict. + if rejection := await self._check_submit_qa_field_gates( + agent_id, task_id, t + ): + return rejection + + # 3. Submit (no catch-up). + submitted = await self.task.submit_qa(agent_id, task_id, notes) + if submitted is not None: + t = submitted + await self._notify_qa(agent_id, task_id, t) + return await self._build_i_am_done_ok(agent_id, task_id, t) + + async def i_am_done_with_catchup( + self, agent_id: UUID, task_id: UUID, notes: str + ) -> Envelope: + """Submit work for QA — opt-in smart catch-up. + + Same tracing-gate preconditions as ``i_am_done``, but auto-runs + the verify / push / PR / submit_qa chain on the dev's behalf + instead of refusing on missing fields. Use this when the dev + explicitly wants the gateway to drive the closure path. + + Pre-gateway behavior: dev had to call each step manually. The + catch-up convenience exists for backward compat with workflows + that rely on the implicit chain. + """ + t = await self.task.get(task_id) + if t is None: + return Envelope.not_found(message=f"task {task_id} not found") + if t.assigned_to != agent_id: + return Envelope.not_authorized( + message="not assigned to you", + remediate="claim it via i_will_work_on(task_id) first", + context_briefing=await self._briefing_for(agent_id, task_id), + ) + if rejection := await self._check_tracing_gates(agent_id, task_id, t): + return rejection + t = await self._run_catch_up(agent_id, task_id, t, notes) + await self._notify_qa(agent_id, task_id, t) + return await self._build_i_am_done_ok(agent_id, task_id, t) + + async def _check_tracing_gates( + self, agent_id: UUID, task_id: UUID, t: Any + ) -> Envelope | None: + """Run progress / reflect / acceptance-criteria tracing gates.""" has_reflect = await self.journal.has_reflect_for_task(agent_id, task_id) gate_ctx = GateContext(journal_reflect_present=has_reflect) gate = check_requirements( @@ -349,16 +405,51 @@ class Choreographer: ], gate_ctx, ) - if not gate.passed: - return await self._build_tracing_gap(agent_id, task_id, gate.missing) + if gate.passed: + return None + return await self._build_tracing_gap(agent_id, task_id, gate.missing) - # 2. Smart catch-up: verification, push, PR, submit_qa - t = await self._run_catch_up(agent_id, task_id, t, notes) + async def _check_submit_qa_field_gates( + self, agent_id: UUID, task_id: UUID, t: Any + ) -> Envelope | None: + """Gate Set E field-level gates restored from tasks.py:903-940 at 0c3d15a. - # 3. Auto-A2A to QA agent for this team - await self._notify_qa(agent_id, task_id, t) + Each missing field becomes its own tracing_gap entry with the + matching pre-gateway error code. + """ + missing: list[str] = [] + hints: list[str] = [] + if not t.self_verified: + missing.append("NOT_SELF_VERIFIED") + hints.append( + "call commit(message=...) to add a self-verified commit first;" + " self_verified is set automatically when you commit on this" + " task's branch" + ) + if not t.commits: + missing.append("NO_COMMITS") + hints.append( + "no commits on this task yet — call commit(message='')" + " before i_am_done" + ) + if t.pr_number is None: + missing.append("NO_PR") + hints.append( + "no PR open — push your branch and open a PR before" + " i_am_done (or call i_am_done_with_catchup to do it auto)" + ) + if not missing: + return None + return Envelope.tracing_gap( + missing=missing, + remediate=" ; ".join(hints), + context_briefing=await self._briefing_for(agent_id, task_id), + ) - # 4. Build evidence for the response + async def _build_i_am_done_ok( + self, agent_id: UUID, task_id: UUID, t: Any + ) -> Envelope: + """Assemble the success envelope for i_am_done / _with_catchup.""" journal_highlights = await self.evidence_repo.journal_highlights_for_task( task_id ) diff --git a/tests/unit/gateway/test_choreographer_dev.py b/tests/unit/gateway/test_choreographer_dev.py index 9c59d30a..0e8b52b6 100644 --- a/tests/unit/gateway/test_choreographer_dev.py +++ b/tests/unit/gateway/test_choreographer_dev.py @@ -263,7 +263,13 @@ async def test_i_have_committed_no_plan_returns_tracing_gap() -> None: @pytest.mark.asyncio -async def test_i_am_done_full_catch_up() -> None: +async def test_i_am_done_with_catchup_full_chain() -> None: + """The catch-up convenience verb auto-runs verify/push/PR/submit_qa. + + Strict ``i_am_done`` requires the dev to have done these steps already + (Gate Set E). When the dev wants the gateway to drive the chain, they + call the explicit catch-up verb. + """ agent_id = uuid4() task_id = uuid4() branch = "feature/backend/abc--def" @@ -379,7 +385,7 @@ async def test_i_am_done_full_catch_up() -> None: deps.evidence_repo.journal_highlights_for_task.return_value = [] c = Choreographer(deps) - env = await c.i_am_done(agent_id, task_id, "all done") + env = await c.i_am_done_with_catchup(agent_id, task_id, "all done") assert env.error is None assert env.status == "awaiting_qa" git_svc.push_branch.assert_awaited_once_with(branch) diff --git a/tests/unit/gateway/test_choreographer_reassignment.py b/tests/unit/gateway/test_choreographer_reassignment.py index e5264f75..fe441a55 100644 --- a/tests/unit/gateway/test_choreographer_reassignment.py +++ b/tests/unit/gateway/test_choreographer_reassignment.py @@ -73,7 +73,8 @@ async def test_i_am_done_reassigns_task_to_qa_agent() -> None: acceptance_criteria_status=[ {"criterion": "AC1", "referencing_artifact_id": "c1"} ], - commits=[], + # Gate Set E requires non-empty commits before submit_qa. + commits=[{"sha": "abc"}], documents=[], dev_notes="", ) @@ -124,7 +125,8 @@ async def test_i_am_done_skips_reassign_when_no_qa_agent() -> None: progress_updates=[{"message": "p"}], acceptance_criteria=[], acceptance_criteria_status=[], - commits=[], + # Gate Set E requires non-empty commits before submit_qa. + commits=[{"sha": "abc"}], documents=[], dev_notes="", ) diff --git a/tests/unit/gateway/test_choreographer_submit_qa_gates.py b/tests/unit/gateway/test_choreographer_submit_qa_gates.py new file mode 100644 index 00000000..409baed8 --- /dev/null +++ b/tests/unit/gateway/test_choreographer_submit_qa_gates.py @@ -0,0 +1,285 @@ +"""Gate Set E: submit-qa field-level gates in Choreographer.i_am_done. + +Pre-gateway location: roboco/api/routes/tasks.py:903-940 (route layer). +The four field-level gates returned 400 errors when the dev tried to +submit for QA without: + +- NOT_SELF_VERIFIED: task.self_verified must be true. +- NO_COMMITS: task.commits must be non-empty. +- NO_PR: task.pr_number must be set. +- NO_PROGRESS: task.progress_updates must have at least one entry. + +The gateway's i_am_done previously called _run_catch_up which silently +auto-ran the full chain. That hid the missing-commits failure mode +(catch-up tried to push nothing, opened an empty PR, etc.). + +Now i_am_done is strict and tells the dev exactly which prerequisite +is missing. A separate i_am_done_with_catchup verb retains the smart- +catch-up behavior for the explicit-opt-in case. +""" + +from __future__ import annotations + +from typing import Any +from unittest.mock import AsyncMock, MagicMock +from uuid import uuid4 + +import pytest +from roboco.services.gateway.choreographer import Choreographer, ChoreographerDeps + + +def _make_deps(**overrides: Any) -> ChoreographerDeps: + base: dict[str, Any] = { + "task": AsyncMock(), + "work_session": AsyncMock(), + "git": AsyncMock(), + "a2a": AsyncMock(), + "journal": AsyncMock(), + "audit": AsyncMock(), + "evidence_repo": AsyncMock(), + } + base.update(overrides) + repo = base["evidence_repo"] + for method in ( + "list_unread_a2a", + "list_unread_mentions", + "list_pending_notifications", + "task_metadata_gaps", + "recent_team_activity", + "blockers_in_lane", + "journal_highlights_for_task", + ): + getattr(repo, method).return_value = [] + return ChoreographerDeps(**base) + + +def _ready_task(task_id: Any, agent_id: Any) -> MagicMock: + """Build a task that satisfies tracing AND field-level gates.""" + return MagicMock( + id=task_id, + status="in_progress", + assigned_to=agent_id, + plan={"x": 1}, + branch_name="feature/backend/abc--def", + work_session_id=uuid4(), + self_verified=True, + pr_number=8, + pr_url="https://x/pr/8", + team="backend", + progress_updates=[{"message": "did x"}], + acceptance_criteria=["AC1"], + acceptance_criteria_status=[ + {"criterion": "AC1", "referencing_artifact_id": "c1"} + ], + commits=[{"sha": "abc"}], + documents=[], + dev_notes="", + ) + + +# --------------------------------------------------------------------------- +# E.1 NOT_SELF_VERIFIED +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_blocks_when_not_self_verified() -> None: + agent_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, agent_id) + t.self_verified = False + task_svc = AsyncMock() + task_svc.get.return_value = t + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + deps = _make_deps(task=task_svc, journal=journal_svc) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "done") + body = env.as_dict() + assert body["error"] == "tracing_gap" + assert "NOT_SELF_VERIFIED" in body["missing"] or "self_verified" in body["missing"] + task_svc.submit_qa.assert_not_awaited() + + +# --------------------------------------------------------------------------- +# E.2 NO_COMMITS +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_blocks_when_no_commits() -> None: + agent_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, agent_id) + t.commits = [] + task_svc = AsyncMock() + task_svc.get.return_value = t + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + deps = _make_deps(task=task_svc, journal=journal_svc) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "done") + body = env.as_dict() + assert body["error"] == "tracing_gap" + assert "NO_COMMITS" in body["missing"] or "commits" in body["missing"] + task_svc.submit_qa.assert_not_awaited() + + +# --------------------------------------------------------------------------- +# E.3 NO_PR +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_blocks_when_no_pr() -> None: + agent_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, agent_id) + t.pr_number = None + task_svc = AsyncMock() + task_svc.get.return_value = t + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + deps = _make_deps(task=task_svc, journal=journal_svc) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "done") + body = env.as_dict() + assert body["error"] == "tracing_gap" + assert "NO_PR" in body["missing"] or "pr_number" in body["missing"] + task_svc.submit_qa.assert_not_awaited() + + +# --------------------------------------------------------------------------- +# E.4 NO_PROGRESS +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_blocks_when_no_progress() -> None: + agent_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, agent_id) + t.progress_updates = [] + task_svc = AsyncMock() + task_svc.get.return_value = t + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + deps = _make_deps(task=task_svc, journal=journal_svc) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "done") + body = env.as_dict() + assert body["error"] == "tracing_gap" + # progress>=1 is the existing tracing_gate Requirement key. + assert "progress>=1" in body["missing"] or "NO_PROGRESS" in body["missing"] + task_svc.submit_qa.assert_not_awaited() + + +# --------------------------------------------------------------------------- +# E.5 happy path: all gates pass → submit_qa runs (NO catch-up) +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_proceeds_when_all_gates_pass() -> None: + agent_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, agent_id) + after_submit = MagicMock( + **{**t.__dict__, "status": "awaiting_qa"}, + ) + task_svc = AsyncMock() + task_svc.get.return_value = t + task_svc.submit_qa.return_value = after_submit + task_svc.qa_agent_for_team.return_value = MagicMock( + id=uuid4(), skills=[{"id": "code_review"}] + ) + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + work_svc = AsyncMock() + work_svc.files_changed.return_value = ["foo.py"] + deps = _make_deps( + task=task_svc, journal=journal_svc, work_session=work_svc + ) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "all done") + body = env.as_dict() + assert body["error"] is None + assert body["status"] == "awaiting_qa" + task_svc.submit_qa.assert_awaited_once() + # Strict path must NOT call submit_verification, push, or create_pr — + # those are catch-up side effects which are now opt-in only. + task_svc.submit_verification.assert_not_awaited() + + +# --------------------------------------------------------------------------- +# E.6 i_am_done_with_catchup retains the smart-catch-up convenience. +# --------------------------------------------------------------------------- + + +@pytest.mark.asyncio +async def test_i_am_done_with_catchup_runs_full_chain() -> None: + agent_id = uuid4() + task_id = uuid4() + initial = _ready_task(task_id, agent_id) + initial.self_verified = False + initial.pr_number = None + after_verify = MagicMock( + **{**initial.__dict__, "self_verified": True, "status": "verifying"} + ) + after_pr_refresh = MagicMock( + **{**after_verify.__dict__, "pr_number": 8, "pr_url": "https://x/pr/8"} + ) + after_submit = MagicMock( + **{**after_pr_refresh.__dict__, "status": "awaiting_qa"} + ) + task_svc = AsyncMock() + task_svc.get.side_effect = [initial, after_pr_refresh] + task_svc.submit_verification.return_value = after_verify + task_svc.submit_qa.return_value = after_submit + task_svc.qa_agent_for_team.return_value = MagicMock( + id=uuid4(), skills=[{"id": "code_review"}] + ) + work_svc = AsyncMock() + work_svc.has_unpushed_commits.return_value = True + work_svc.files_changed.return_value = ["foo.py"] + git_svc = AsyncMock() + git_svc.create_pr.return_value = {"pr_number": 8, "pr_url": "https://x/pr/8"} + journal_svc = AsyncMock() + journal_svc.has_reflect_for_task.return_value = True + deps = _make_deps( + task=task_svc, + journal=journal_svc, + work_session=work_svc, + git=git_svc, + ) + c = Choreographer(deps) + + env = await c.i_am_done_with_catchup(agent_id, task_id, "all done") + body = env.as_dict() + assert body["error"] is None + assert body["status"] == "awaiting_qa" + task_svc.submit_verification.assert_awaited_once() + git_svc.push_branch.assert_awaited_once() + git_svc.create_pr.assert_awaited_once() + + +@pytest.mark.asyncio +async def test_i_am_done_blocks_unauthorized() -> None: + """Existing not_authorized check still applies.""" + agent_id = uuid4() + other_id = uuid4() + task_id = uuid4() + t = _ready_task(task_id, other_id) + task_svc = AsyncMock() + task_svc.get.return_value = t + deps = _make_deps(task=task_svc) + c = Choreographer(deps) + + env = await c.i_am_done(agent_id, task_id, "done") + body = env.as_dict() + assert body["error"] == "not_authorized"