feat(gateway): carry intake technical depth down the chain + widen review coherence scope (#491)

Two structural issues flagged by the CEO:

1. Task technical-depth dilution — intake's rich analysis (file:line
   targets, code examples, rationale) was getting lost as it traveled
   umbrella -> root-subtask -> cell -> dev. The detail IS preserved in
   Task.description; the dilution was in delegation (PMs re-authoring)
   and the intake prompt not demanding depth.

   Fixes:
   - evidence_repo: ancestor_context_for_task walks the parent chain
     (cycle-guarded, depth-capped 16, desc-clipped 1500) and surfaces it
     as parent_context in the evidence payload, so a leaf dev finally
     sees the upstream intake analysis instead of a bare title.
   - evidence_builder: Task.description now rides in the payload;
     EvidencePayload gains description + parent_context (omit-when-empty
     so no null noise).
   - orchestrator: _description_body (capped 4000) injects the
     description into the dev spawn prompt + SessionStart briefing.
   - role prompts (main_pm/cell_pm/developer/prompter): teach pass-the-
     torch, don't-dim-it; prompter now demands file:line/code-examples
     in the_work/notes (reconciled with the no-code-level-ACs-on-roots
     rule). main_pm's brief-not-a-spec scoped: not-a-spec applies to the
     solution only, facts forward verbatim.

2. PR-review/QA scope too narrow — they only checked the AC checklist,
   not whether the change is coherent with project structure/intent.

   Fixes:
   - qa.md + pr_reviewer.md: Coherence & intent rule (intent via
     description+parent_context, coherence with project patterns,
     standards). Criterion-less major findings allowed for intent drift
     (Finding.criterion is optional).
   - parent_context + description wired into the gate/QA/inbound-PR
     evidence builders (fail-open, logged).

Skipped per YAGNI: a technical_spec JSONB column (detail is already in
description) and a criterion_kind enum (criterion is already optional).

All gates green: ruff, mypy (1152), pytest (12883 passed, 94.82% cov),
xenon, vulture, bandit, pip-audit, deptry, alembic, import-linter,
foundation-check.

Co-authored-by: Renn F <rennf93@users.noreply.github.com>
This commit is contained in:
Renzo F
2026-07-13 08:06:57 +02:00
committed by GitHub
co-authored by Renn F
parent 192524265c
commit ba7135ba50
19 changed files with 536 additions and 34 deletions
+146 -1
View File
@@ -13,7 +13,25 @@ from unittest.mock import AsyncMock, MagicMock
from uuid import uuid4
import pytest
from roboco.services.gateway.evidence_repo import EvidenceRepo
from roboco.services.gateway.evidence_repo import (
_ANCESTOR_DESC_CAP,
_HIERARCHY_DEPTH_CAP,
EvidenceRepo,
)
def _scalar_result(value: object) -> MagicMock:
"""A query result whose ``scalar_one_or_none()`` returns ``value``."""
r = MagicMock()
r.scalar_one_or_none.return_value = value
return r
def _rows_result(rows: list[object]) -> MagicMock:
"""A query result whose ``all()`` returns ``rows``."""
r = MagicMock()
r.all.return_value = rows
return r
def _empty_repo() -> EvidenceRepo:
@@ -233,3 +251,130 @@ async def test_journal_highlights_for_task_maps_rows_with_author() -> None:
"timestamp": ts.isoformat(),
}
]
# ---------------------------------------------------------------------------
# Parent-chain walk + ancestor context (the intake-analysis torch carrier).
# ---------------------------------------------------------------------------
@pytest.mark.asyncio
async def test_ancestor_task_ids_walks_parent_to_root() -> None:
leaf, parent, grand = uuid4(), uuid4(), uuid4()
db = MagicMock()
db.execute = AsyncMock(
side_effect=[
_scalar_result(parent),
_scalar_result(grand),
_scalar_result(None), # grand is a root
]
)
repo = EvidenceRepo(db)
assert await repo._ancestor_task_ids(leaf) == [parent, grand]
@pytest.mark.asyncio
async def test_ancestor_task_ids_cycle_guard_stops() -> None:
leaf, parent = uuid4(), uuid4()
db = MagicMock()
# leaf -> parent -> leaf (cycles back to the start, already in `seen`).
db.execute = AsyncMock(side_effect=[_scalar_result(parent), _scalar_result(leaf)])
repo = EvidenceRepo(db)
assert await repo._ancestor_task_ids(leaf) == [parent]
@pytest.mark.asyncio
async def test_ancestor_task_ids_missing_row_returns_empty() -> None:
db = MagicMock()
db.execute = AsyncMock(side_effect=[_scalar_result(None)])
repo = EvidenceRepo(db)
assert await repo._ancestor_task_ids(uuid4()) == []
@pytest.mark.asyncio
async def test_ancestor_task_ids_depth_capped() -> None:
start = uuid4()
chain = [uuid4() for _ in range(_HIERARCHY_DEPTH_CAP + 5)]
db = MagicMock()
db.execute = AsyncMock(side_effect=[_scalar_result(p) for p in chain])
repo = EvidenceRepo(db)
result = await repo._ancestor_task_ids(start)
assert len(result) == _HIERARCHY_DEPTH_CAP
assert result == chain[:_HIERARCHY_DEPTH_CAP]
@pytest.mark.asyncio
async def test_ancestor_context_for_task_parentless_returns_empty() -> None:
db = MagicMock()
db.execute = AsyncMock(side_effect=[_scalar_result(None)])
repo = EvidenceRepo(db)
assert await repo.ancestor_context_for_task(uuid4()) == []
@pytest.mark.asyncio
async def test_ancestor_context_for_task_maps_chain_with_depth() -> None:
leaf, parent, grand = uuid4(), uuid4(), uuid4()
row_p = SimpleNamespace(id=parent, title="Cell PM slice", description="p-desc")
row_g = SimpleNamespace(id=grand, title="Root", description="g-desc")
db = MagicMock()
db.execute = AsyncMock(
side_effect=[
_scalar_result(parent),
_scalar_result(grand),
_scalar_result(None), # end of _ancestor_task_ids
_rows_result([row_p, row_g]), # the batch fetch
]
)
repo = EvidenceRepo(db)
assert await repo.ancestor_context_for_task(leaf) == [
{
"task_id": str(parent),
"depth": 1,
"title": "Cell PM slice",
"description": "p-desc",
},
{"task_id": str(grand), "depth": 2, "title": "Root", "description": "g-desc"},
]
@pytest.mark.asyncio
async def test_ancestor_context_for_task_skips_missing_ancestor_row() -> None:
leaf, parent, grand = uuid4(), uuid4(), uuid4()
row_p = SimpleNamespace(id=parent, title="Cell PM slice", description="p-desc")
db = MagicMock()
db.execute = AsyncMock(
side_effect=[
_scalar_result(parent),
_scalar_result(grand),
_scalar_result(None),
_rows_result([row_p]), # grand's row is missing from the batch
]
)
repo = EvidenceRepo(db)
assert await repo.ancestor_context_for_task(leaf) == [
{
"task_id": str(parent),
"depth": 1,
"title": "Cell PM slice",
"description": "p-desc",
},
]
@pytest.mark.asyncio
async def test_ancestor_context_for_task_clips_long_description() -> None:
leaf, parent = uuid4(), uuid4()
row_p = SimpleNamespace(
id=parent, title="P", description="x" * (_ANCESTOR_DESC_CAP + 500)
)
db = MagicMock()
db.execute = AsyncMock(
side_effect=[
_scalar_result(parent),
_scalar_result(None),
_rows_result([row_p]),
]
)
repo = EvidenceRepo(db)
ctx = await repo.ancestor_context_for_task(leaf)
assert len(ctx[0]["description"]) == _ANCESTOR_DESC_CAP