feat(auditor): waive_finding verb + findings queue panel

Wire the long-unwired mark_waived repo method to a new auditor-only
flow verb waive_finding, severity-scoped to minor/nit (blocker/major
must be fixed, never waived), requiring a note, with a task.finding_waived
audit event and no task status change. Add the verb to the IntentSpec
table (auto-derived into the auditor manifest), the flow_auditor route,
and the flow_server MCP tool.

Surface open review findings (cross-task, blocking-first) on the
auditor dashboard via ReviewFindingsRepository.list_open_findings and a
new findings field on AuditorDashboard. Restore the panel's 4-card
auditor layout with a new read-only FindingsQueuePanel as the 4th card.
This commit is contained in:
Renn F
2026-07-14 08:56:55 +02:00
committed by Renzo F
parent 62e19ea729
commit 05a83f45cb
18 changed files with 700 additions and 8 deletions
+2 -2
View File
@@ -214,7 +214,7 @@ backlog -> pending -> claimed -> in_progress -> [blocked|paused] -> verifying
**Revision findings ledger (always-on, no flag — core lifecycle).** QA/PR-gate/PM/CEO bounce feedback used to be prose-only: `issues: list[str]` flattened into free text with no structural anchor, `request_changes`/`ceo_reject` had no structured note at all, and the raw `dev_notes` append both used was silently overwritten by the very next `note(scope='handoff')` call — a live data-loss bug. `fail_review` (QA), `pr_fail` (in-path PR gate), `request_changes` (PM merge reject), and `ceo_reject` all now take structured `findings: list[dict]` — validated into `Finding` (`file` repo-relative ≤300 chars/no `..`, `line` ≥1, `severity` blocker\|major\|minor\|nit, `criterion` must match an AC id or its exact text, `expected`/`actual` ≤300, `fix` ≤500, `evidence` ≤2000) — with a soft nudge above 5 findings and a hard reject above 10 in one call (`roboco/services/gateway/choreographer/findings.py`). `issues=[...]` still works this release as a shim (each string → a file-less `severity=major` finding, deprecation-logged) and merges with `findings` rather than one silently dropping the other. Every producer inserts one append-only row per finding into `task_review_findings` (migration 071; `origin` qa\|pr_gate\|pm\|ceo, `round` = `revision_count+1` read pre-transition, `status` open→addressed→verified\|waived) via `ReviewFindingsRepository`, then writes a structured note whose `summary` IS the deterministic per-finding rendering `[F-id8] file:line (severity) — expected → actual → fix`, mirrored into `qa_notes`/`pr_reviewer_notes`/the new `pm_notes` column (new `PmReviewContent` "pm_review" content type). `ceo_reject` now validates its reason (previously could 500 on an empty/trivial one) and stamps it as one `origin=ceo` `blocker` finding; on a branchless coordination root — which routes to `pending` via `admin_set_status`, bypassing the normal audit chokepoint — it bumps `revision_count` and emits `task.ceo_reject` directly instead of silently skipping both. New audit events `task.request_changes`/`task.ceo_reject` join `task.qa_fail`/`task.pr_fail` in `_audit_events_for` so rework metrics attribute every bounce kind, not just QA/PR-gate.
Resolution: `i_am_done`/`submit_up`/`submit_root` all gain `resolved_findings` (`{finding_id, commit?, note?}`), gated by a new `Requirement.FINDINGS_ADDRESSED` — every OPEN finding on the task must be named (a fuzzy 8-char-prefix match against `[F-id8]`) or the envelope rejects, listing the still-open ids. `pass_review`/`pr_pass`/`complete` bulk-verify their own origin's `addressed` findings same-transaction (a stamp failure fails the verb outright, not best-effort); `ceo_approve` stamps `ceo`-origin best-effort. `mark_waived` exists on the repository but no verb calls it yet — a deliberate unwired follow-up.
Resolution: `i_am_done`/`submit_up`/`submit_root` all gain `resolved_findings` (`{finding_id, commit?, note?}`), gated by a new `Requirement.FINDINGS_ADDRESSED` — every OPEN finding on the task must be named (a fuzzy 8-char-prefix match against `[F-id8]`) or the envelope rejects, listing the still-open ids. `pass_review`/`pr_pass`/`complete` bulk-verify their own origin's `addressed` findings same-transaction (a stamp failure fails the verb outright, not best-effort); `ceo_approve` stamps `ceo`-origin best-effort. `mark_waived` is wired to the auditor-only `waive_finding` flow verb (severity-scoped: blocker/major must be fixed, never waived; only minor/nit open findings are waivable, with a required note and a `task.finding_waived` audit event; no task status change).
Delivery: `evidence()`/`build_task_handoff` carry `revision_findings` (open only, capped) so a bounced dev finally gets what `developer.md` promises instead of nothing; `claim_review`/`claim_gate_review` additionally carry `prior_findings` (the full ledger) so a round-2+ reviewer checks prior findings instead of re-deriving them blind. The orchestrator's `REVISION_REQUIRED` dev prompt and the PM triage "bounced" block render open findings inline with the same rendering; A2A fail bodies share it. `GET /api/tasks/{id}/findings` (capped 500, SQL-aggregated per-origin/status summary + `total`/`truncated`) backs the panel's task-detail Findings tab and a `bounced xN` header chip (`revision_count`); metrics attribute `pm_rejects`/`ceo_rejects` + open/total findings counts per task; vault task notes render a capped `## Findings` section (fail-open fetch, never blocks the note write).
@@ -349,7 +349,7 @@ Each agent gets a **spawn manifest** at `/app/tool-manifest.json` listing the ve
| pr_reviewer | `give_me_work`, `claim_pr_review`, `post_pr_review` (inbound external/fork PRs), `claim_gate_review`, `pr_pass`, `pr_fail` (in-path assembled-PR gate), `unclaim` |
| product_owner | `triage`, `escalate_to_ceo` |
| head_marketing| `triage`, `escalate_to_ceo` |
| auditor | `triage` (read-only — no `dm`) |
| auditor | `triage`, `waive_finding` (read-only — no `dm`) |
| prompter | (none beyond `i_am_idle` — not a delivery-lifecycle role; intake interviewer, human-only) |
| secretary | (none beyond `i_am_idle` — human-only chief-of-staff; reads company state + runs gated CEO directives) |
@@ -10,6 +10,7 @@ import {
import { QualityMetricsPanel } from "./quality-metrics-panel";
import { FlaggedItemsPanel } from "./flagged-items-panel";
import { ReportsPanel } from "./reports-panel";
import { FindingsQueuePanel } from "./findings-queue-panel";
import { Button } from "@/components/ui/button";
import { FileText } from "lucide-react";
import { toast } from "sonner";
@@ -73,11 +74,17 @@ export function AuditorDashboard() {
</div>
</div>
{/* Quality Metrics */}
{/* Top Row: Open Findings + Quality Metrics */}
<div className="grid grid-cols-1 lg:grid-cols-2 xl:grid-cols-2 2xl:grid-cols-2 gap-6">
<FindingsQueuePanel
findings={dashboard?.findings}
isLoading={loadingDashboard}
/>
<QualityMetricsPanel
metrics={dashboard?.metrics}
isLoading={loadingDashboard}
/>
</div>
{/* Bottom Row: Flagged Items + Reports */}
<div className="grid grid-cols-1 lg:grid-cols-2 xl:grid-cols-2 2xl:grid-cols-2 gap-6">
@@ -0,0 +1,120 @@
"use client";
import { AuditorFinding } from "@/types";
import { Card, CardContent, CardHeader, CardTitle } from "@/components/ui/card";
import { Badge } from "@/components/ui/badge";
import { Skeleton } from "@/components/ui/skeleton";
import { ScrollArea } from "@/components/ui/scroll-area";
import { ListChecks } from "lucide-react";
import Link from "next/link";
interface FindingsQueuePanelProps {
findings: AuditorFinding[] | undefined;
isLoading: boolean;
}
// Severity stored as the finding's lowercase value (blocker/major/minor/nit).
const severityColors: Record<string, string> = {
blocker: "bg-red-100 text-red-700",
major: "bg-orange-100 text-orange-700",
minor: "bg-yellow-100 text-yellow-700",
nit: "bg-blue-100 text-blue-700",
};
const severityOrder: Record<string, number> = {
blocker: 0,
major: 1,
minor: 2,
nit: 3,
};
export function FindingsQueuePanel({
findings,
isLoading,
}: FindingsQueuePanelProps) {
// The API already returns blocking-first, but keep the sort stable client-side
// in case a later re-fetch reorders.
const sorted = [...(findings ?? [])].sort((a, b) => {
const diff = (severityOrder[a.severity] ?? 9) - (severityOrder[b.severity] ?? 9);
if (diff !== 0) return diff;
return (b.created_at ?? "").localeCompare(a.created_at ?? "");
});
return (
<Card>
<CardHeader className="pb-3">
<div className="flex items-center justify-between">
<div className="flex items-center gap-2">
<CardTitle className="text-lg flex items-center gap-2">
<ListChecks className="h-5 w-5" />
Open Findings
</CardTitle>
{sorted.length > 0 && (
<Badge variant="destructive">{sorted.length}</Badge>
)}
</div>
</div>
</CardHeader>
<CardContent>
{isLoading ? (
<div className="space-y-3">
{[...Array(3)].map((_, i) => (
<Skeleton key={i} className="h-24" />
))}
</div>
) : sorted.length === 0 ? (
<div className="text-center py-8 text-muted-foreground text-sm">
<ListChecks className="h-8 w-8 mx-auto mb-2 opacity-50" />
No open review findings
</div>
) : (
<ScrollArea className="h-[400px] pr-4">
<div className="space-y-3">
{sorted.map((finding) => (
<div
key={finding.id}
className="p-4 rounded-lg border bg-muted/50"
>
<div className="flex items-center gap-2 mb-1 flex-wrap">
<Badge
className={
(severityColors[finding.severity] ?? "") + " text-xs"
}
>
{finding.severity}
</Badge>
<Badge variant="outline" className="text-xs">
{finding.origin}
</Badge>
<span className="text-xs text-muted-foreground">
round {finding.round}
</span>
</div>
<p className="text-sm text-muted-foreground mb-2">
{finding.actual ?? finding.expected ?? finding.criterion ?? "—"}
</p>
<div className="flex items-center gap-4 text-xs text-muted-foreground">
{finding.file && (
<span className="font-mono">
{finding.file}
{finding.line ? `:${finding.line}` : ""}
</span>
)}
<Link
href={"/tasks/" + finding.task_id}
prefetch={false}
>
<span className="text-primary hover:underline">
Task #{finding.task_id.slice(0, 8)}
</span>
</Link>
</div>
</div>
))}
</div>
</ScrollArea>
)}
</CardContent>
</Card>
);
}
+1
View File
@@ -1187,6 +1187,7 @@ export const mockAuditorDashboard = {
},
],
recent_reports: mockAuditorReports,
findings: [],
};
// Mock mode: dev = mock, production = real backend
+15
View File
@@ -611,6 +611,20 @@ export interface AuditorReport {
sent_at: string | null;
}
export interface AuditorFinding {
id: string;
task_id: string;
origin: string;
severity: string;
file: string | null;
line: number | null;
criterion: string | null;
expected: string | null;
actual: string | null;
round: number;
created_at: string | null;
}
export interface AuditorDashboard {
flagged_items: AuditorFlag[];
metrics: Record<string, number>;
@@ -621,6 +635,7 @@ export interface AuditorDashboard {
team: string | null;
}>;
recent_reports: AuditorReport[];
findings: AuditorFinding[];
}
export interface TeamHealth {
+1
View File
@@ -117,6 +117,7 @@ async def get_auditor_dashboard(
metrics=metrics,
audit_queue=audit_queue,
recent_reports=recent_reports,
findings=await service.get_open_findings(limit=20),
)
+19 -1
View File
@@ -11,7 +11,11 @@ from guard_core.handlers.behavior_handler import BehaviorRule
from roboco.api.deps import get_choreographer
from roboco.api.routes.v1._role_dep import envelope_to_response, require_auditor
from roboco.api.schemas.v1.flow import IAmIdleRequest, TriageRequest
from roboco.api.schemas.v1.flow import (
IAmIdleRequest,
TriageRequest,
WaiveFindingRequest,
)
from roboco.security import guard_deco
from roboco.services.gateway.choreographer import Choreographer
@@ -44,6 +48,20 @@ async def triage(
return envelope_to_response(env, request)
@router.post("/waive_finding")
@guard_deco.rate_limit(requests=30, window=60)
@guard_deco.content_type_filter(["application/json"])
@guard_deco.behavior_analysis(_RUNAWAY_RULES)
async def waive_finding(
request: Request,
body: WaiveFindingRequest,
x_agent_id: _AgentIdHeader,
choreographer: _ChoreographerDep,
) -> dict:
env = await choreographer.waive_finding(x_agent_id, body.finding_id, body.note)
return envelope_to_response(env, request)
@router.post("/i_am_idle")
@guard_deco.rate_limit(requests=30, window=60)
@guard_deco.content_type_filter(["application/json"])
+1
View File
@@ -54,6 +54,7 @@ class AuditorDashboard(BaseModel):
metrics: dict[str, Any]
audit_queue: list[dict[str, Any]]
recent_reports: list[AuditorReport]
findings: list[dict[str, Any]] = Field(default_factory=list)
class TeamHealth(BaseModel):
+12
View File
@@ -284,6 +284,18 @@ class TriageRequest(BaseModel):
"""Empty request body."""
class WaiveFindingRequest(BaseModel):
finding_id: UUID
note: str = Field(
...,
min_length=1,
description=(
"Why this finding is waived rather than fixed — recorded on the "
"ledger row and in audit. Only minor/nit findings are waivable."
),
)
class UnblockRequest(BaseModel):
task_id: UUID
reason: str = Field(
+13
View File
@@ -1591,6 +1591,19 @@ _INTENT_VERBS: dict[str, IntentSpec] = {
side_effects=(),
next_hint=lambda _t: "act on a listed task or i_am_idle",
),
"waive_finding": IntentSpec(
name="waive_finding",
allowed_roles=frozenset({Role.AUDITOR}),
description=(
"Waive one minor/nit review finding by id with a required note. "
"Blocker/major findings must be fixed, never waived. No task "
"status change."
),
composes=(),
extra_preconditions=(),
side_effects=(),
next_hint=lambda _t: "finding waived; triage() for next item or i_am_idle",
),
}
+14
View File
@@ -879,6 +879,18 @@ def escalate_to_ceo(task_id: str, reason: str) -> dict[str, Any]:
return _post(_role_path("escalate_to_ceo"), {"task_id": task_id, "reason": reason})
def waive_finding(finding_id: str, note: str) -> dict[str, Any]:
"""Auditor: waive one minor/nit review finding by id with a required note.
Blocker/major findings must be fixed, never waived. The finding id is the
``[F-<id8>]`` prefix shown in task notes / triage. No task status changes.
"""
return _post(
_role_path("waive_finding"),
{"finding_id": finding_id, "note": note},
)
# ---------- Cell PM + Main PM extras ----------
# i_will_plan, delegate, submit_up, give_me_work — restore the pre-Phase-4
# PM lifecycle so PMs can drive parent tasks instead of stalling.
@@ -1108,6 +1120,8 @@ _TOOLS: dict[str, Any] = {
"declare_coverage": declare_coverage,
# board / main pm
"escalate_to_ceo": escalate_to_ceo,
# auditor
"waive_finding": waive_finding,
}
+32
View File
@@ -319,6 +319,38 @@ class DashboardService(BaseService):
"active_blockers": blockers.active_blockers,
}
async def get_open_findings(self, limit: int = 20) -> list[dict[str, Any]]:
"""Open review findings across all tasks for the auditor's findings queue.
Blocking severity first then newest. Fails open (empty list) the
dashboard must never break on a findings-query error.
"""
from roboco.services.repositories.review_findings import (
ReviewFindingsRepository,
)
try:
repo = ReviewFindingsRepository(self.session)
rows = await repo.list_open_findings(limit=limit)
except Exception:
return []
return [
{
"id": str(r.id),
"task_id": str(r.task_id),
"origin": r.origin,
"severity": r.severity,
"file": r.file,
"line": r.line,
"criterion": r.criterion,
"expected": r.expected,
"actual": r.actual,
"round": r.round,
"created_at": r.created_at.isoformat() if r.created_at else None,
}
for r in rows
]
async def get_all_agent_status(self, team: Team | None = None) -> dict[str, Any]:
"""Return agent-status summary (counts + per-agent snapshot)."""
query = select(AgentTable)
@@ -14,9 +14,15 @@ the actual class is composed in ``__init__.py`` and inherits from
from __future__ import annotations
import contextlib
from typing import TYPE_CHECKING
from roboco.foundation.policy.content import Severity
from roboco.services.gateway.envelope import Envelope
from roboco.services.repositories.review_findings import (
STATUS_OPEN,
ReviewFindingsRepository,
)
if TYPE_CHECKING:
from uuid import UUID
@@ -27,6 +33,9 @@ if TYPE_CHECKING:
else:
_Base = object
# Severity stored as the enum's string value (``f.severity.value`` at insert).
_BLOCKING = frozenset({Severity.BLOCKER.value, Severity.MAJOR.value})
class BoardMixin(_Base):
"""Board (Product Owner + Head Marketing) + Auditor verbs."""
@@ -88,3 +97,76 @@ class BoardMixin(_Base):
next="no anomalies — i_am_idle",
context_briefing=await self._briefing_for(auditor_agent_id, None),
)
async def waive_finding(
self, auditor_agent_id: UUID, finding_id: UUID, note: str
) -> Envelope:
"""Waive one minor/nit finding by id with a required note.
The auditor is the only role that can close a finding without a dev
fix but only for non-blocking severity (minor/nit). Blocker/major
must be fixed, never waived. No task status change: the ledger row
moves ``open -> waived`` and an audit event records the decision.
``mark_waived`` is the long-unwired repo method this finally calls.
"""
repo = ReviewFindingsRepository(self.task.session)
row = await repo.get(finding_id)
if row is None:
return Envelope.not_found(
message=f"finding {str(finding_id)[:8]} not found.",
remediate=(
"find open findings via the task's GET /findings or triage()."
),
)
if row.status != STATUS_OPEN:
return Envelope.invalid_state(
message=(
f"finding {str(finding_id)[:8]} is already {row.status}; "
"only open findings can be waived."
),
remediate=(
"pick an open finding — waived/addressed/verified "
"rows are immutable."
),
)
if row.severity in _BLOCKING:
return Envelope.invalid_state(
message=(
f"finding {str(finding_id)[:8]} is {row.severity}"
"blocker/major findings must be fixed, never waived."
),
remediate=(
"leave it for the dev to address; waive only minor/nit findings."
),
)
clean_note = note.strip()
if not clean_note:
return Envelope.invalid_state(
message=(
"a waive requires a note explaining why this finding "
"is not worth a fix."
),
remediate="pass note=<why this minor/nit is waived>.",
)
await repo.mark_waived(finding_id, clean_note)
with contextlib.suppress(Exception):
await self.audit.log_task_event(
event_type="task.finding_waived",
task_id=row.task_id,
agent_id=auditor_agent_id,
severity="info",
details={
"finding_id": str(finding_id),
"severity": row.severity,
"origin": row.origin,
"note": clean_note[:300],
},
)
return Envelope.ok(
status="waived",
task_id=str(row.task_id),
next=(
f"finding {str(finding_id)[:8]} waived; triage() for next item "
"or i_am_idle"
),
)
@@ -162,3 +162,25 @@ class ReviewFindingsRepository(BaseRepository[TaskReviewFindingTable]):
row.resolution_note = note[:_RESOLUTION_NOTE_CAP]
await self.session.flush()
return True
async def list_open_findings(
self, *, limit: int = 20
) -> list[TaskReviewFindingTable]:
"""Cross-task open findings, blocking severity first then newest.
For the auditor dashboard's findings queue — the backlog of unresolved
review findings the auditor can triage / waive. Capped (the dashboard
is a glance, not a ledger view ``GET /api/tasks/{id}/findings`` is
the full per-task ledger).
"""
stmt = (
select(TaskReviewFindingTable)
.where(TaskReviewFindingTable.status == STATUS_OPEN)
.order_by(
TaskReviewFindingTable.severity.asc(),
TaskReviewFindingTable.created_at.desc(),
)
.limit(limit)
)
result = await self.session.execute(stmt)
return list(result.scalars().all())
@@ -0,0 +1,35 @@
"""IntentSpec for the auditor ``waive_finding`` verb.
``mark_waived`` sat unwired on the repository since the findings ledger
landed (PR #486) — a deliberate follow-up. The auditor is the role that
can close a finding without a dev fix, but only for non-blocking severity.
This pins the spec: auditor-only, severity-scoped at the verb body (the
IntentSpec carries no task precondition ``composes=()`` like ``triage``).
"""
from __future__ import annotations
from roboco.foundation.identity import Role
from roboco.foundation.policy.lifecycle import intents_for_role
from roboco.services.gateway.role_config import _AUDITOR_FLOW
def test_waive_finding_is_an_auditor_flow_verb() -> None:
assert "waive_finding" in intents_for_role(Role.AUDITOR)
assert "waive_finding" in _AUDITOR_FLOW
def test_waive_finding_is_auditor_only() -> None:
for role in (
Role.DEVELOPER,
Role.QA,
Role.DOCUMENTER,
Role.CELL_PM,
Role.MAIN_PM,
Role.PR_REVIEWER,
Role.PRODUCT_OWNER,
Role.HEAD_MARKETING,
):
assert "waive_finding" not in intents_for_role(role), (
f"{role} must not get waive_finding"
)
@@ -0,0 +1,169 @@
"""Choreographer.waive_finding — the auditor's close-without-fix verb.
``mark_waived`` was the long-unwired repo method; this is its only caller.
Severity-scoped: blocker/major must be fixed, never waived. Only open
findings are waivable, and a note is required. No task status changes
the ledger row ``open -> waived`` plus a ``task.finding_waived`` audit
event is the durable record.
"""
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() -> ChoreographerDeps:
base = {
"task": AsyncMock(),
"work_session": AsyncMock(),
"git": AsyncMock(),
"a2a": AsyncMock(),
"journal": AsyncMock(),
"audit": AsyncMock(),
"evidence_repo": AsyncMock(),
}
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 _finding_row(
*,
severity: str = "minor",
status: str = "open",
origin: str = "qa",
) -> MagicMock:
row = MagicMock()
row.id = uuid4()
row.task_id = uuid4()
row.severity = severity
row.status = status
row.origin = origin
return row
def _patch_repo(monkeypatch: pytest.MonkeyPatch, row: Any | None) -> MagicMock:
"""Patch the board-module ReviewFindingsRepository to return ``row`` from
``get`` and a recording ``mark_waived``. ``row=None`` simulates not-found."""
repo_mock = MagicMock()
repo_mock.get = AsyncMock(return_value=row)
repo_mock.mark_waived = AsyncMock(return_value=True)
monkeypatch.setattr(
"roboco.services.gateway.choreographer.board.ReviewFindingsRepository",
lambda *_a, **_k: repo_mock,
)
return repo_mock
def _choreographer(monkeypatch: pytest.MonkeyPatch, row: Any | None):
deps = _make_deps()
deps.task.session = MagicMock()
c = Choreographer(deps)
return c, _patch_repo(monkeypatch, row)
@pytest.mark.asyncio
async def test_waive_minor_open_finding_succeeds(
monkeypatch: pytest.MonkeyPatch,
) -> None:
row = _finding_row(severity="minor", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, "cosmetic, not worth a fix")
assert env.error is None
assert env.status == "waived"
repo_mock.mark_waived.assert_awaited_once_with(row.id, "cosmetic, not worth a fix")
c.audit.log_task_event.assert_awaited_once()
assert c.audit.log_task_event.await_args.kwargs["event_type"] == (
"task.finding_waived"
)
@pytest.mark.asyncio
async def test_waive_nit_open_finding_succeeds(
monkeypatch: pytest.MonkeyPatch,
) -> None:
row = _finding_row(severity="nit", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, "preference only")
assert env.error is None
repo_mock.mark_waived.assert_awaited_once()
@pytest.mark.asyncio
async def test_waive_rejects_blocker(monkeypatch: pytest.MonkeyPatch) -> None:
row = _finding_row(severity="blocker", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, "try to skip the fix")
assert env.error == "invalid_state"
repo_mock.mark_waived.assert_not_awaited()
@pytest.mark.asyncio
async def test_waive_rejects_major(monkeypatch: pytest.MonkeyPatch) -> None:
row = _finding_row(severity="major", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, "try to skip the fix")
assert env.error == "invalid_state"
repo_mock.mark_waived.assert_not_awaited()
@pytest.mark.asyncio
async def test_waive_rejects_already_addressed(
monkeypatch: pytest.MonkeyPatch,
) -> None:
row = _finding_row(severity="minor", status="addressed")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, "already closed")
assert env.error == "invalid_state"
repo_mock.mark_waived.assert_not_awaited()
@pytest.mark.asyncio
async def test_waive_rejects_blank_note(
monkeypatch: pytest.MonkeyPatch,
) -> None:
row = _finding_row(severity="minor", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
env = await c.waive_finding(uuid4(), row.id, " ")
assert env.error == "invalid_state"
repo_mock.mark_waived.assert_not_awaited()
@pytest.mark.asyncio
async def test_waive_unknown_finding_returns_not_found(
monkeypatch: pytest.MonkeyPatch,
) -> None:
c, repo_mock = _choreographer(monkeypatch, None)
env = await c.waive_finding(uuid4(), uuid4(), "note")
assert env.error == "not_found"
repo_mock.mark_waived.assert_not_awaited()
@pytest.mark.asyncio
async def test_waive_succeeds_even_if_audit_log_fails(
monkeypatch: pytest.MonkeyPatch,
) -> None:
"""The audit event is best-effort: a log failure must not undo the waive."""
row = _finding_row(severity="minor", status="open")
c, repo_mock = _choreographer(monkeypatch, row)
c.audit.log_task_event = AsyncMock(side_effect=RuntimeError("db gone"))
env = await c.waive_finding(uuid4(), row.id, "still waive me")
assert env.error is None
repo_mock.mark_waived.assert_awaited_once()
@@ -0,0 +1,98 @@
"""flow_server exposes the auditor's ``waive_finding`` tool.
The verb is auto-derived into the auditor manifest via
``intents_for_role(Role.AUDITOR)``; this pins that the MCP layer registers
it under the public name and POSTs the right payload to the auditor path.
"""
from __future__ import annotations
import importlib
import json
from typing import TYPE_CHECKING, Any
from unittest.mock import MagicMock, patch
import pytest
from roboco.foundation.identity import Role
from roboco.foundation.policy.lifecycle import intents_for_role
if TYPE_CHECKING:
import types
from pathlib import Path
def _auditor_manifest() -> dict[str, object]:
return {
"agent_id": "00000000-0000-0000-0000-000000000004",
"role": "auditor",
"team": "board",
"workspace_path": "/tmp/test",
"flow_tools": list(intents_for_role(Role.AUDITOR)),
"do_tools": [],
"read_tools": [],
"write_tools": [],
"bash_allowed": True,
"subagent_allowed": False,
"subagent_model": None,
"env": {},
}
@pytest.fixture()
def flow_module_auditor(
monkeypatch: pytest.MonkeyPatch, tmp_path: Path
) -> types.ModuleType:
manifest_path = tmp_path / "tool-manifest.json"
manifest_path.write_text(json.dumps(_auditor_manifest()))
monkeypatch.setenv("ROBOCO_AGENT_ID", "00000000-0000-0000-0000-000000000004")
monkeypatch.setenv("ROBOCO_AGENT_ROLE", "auditor")
monkeypatch.setenv("ROBOCO_ORCHESTRATOR_URL", "http://test-orchestrator:8000")
monkeypatch.setenv("ROBOCO_SDK_URL", "http://test-sdk:9000")
monkeypatch.setenv("ROBOCO_TOOL_MANIFEST_PATH", str(manifest_path))
import roboco.mcp.flow_server as srv
importlib.reload(srv)
return srv
def test_waive_finding_registers_for_auditor_manifest(
flow_module_auditor: types.ModuleType,
) -> None:
registered = flow_module_auditor._register_tools()
assert "waive_finding" in registered, (
f"waive_finding not registered for auditor manifest. "
f"Registered: {sorted(registered)}"
)
def test_waive_finding_posts_to_auditor_path(
flow_module_auditor: types.ModuleType,
) -> None:
captured: list[tuple[str, Any]] = []
def _client_factory(*_a: object, **_kw: object) -> MagicMock:
client = MagicMock()
client.__enter__ = MagicMock(return_value=client)
client.__exit__ = MagicMock(return_value=False)
def _post(url: str, **kwargs: object) -> MagicMock:
captured.append((url, kwargs.get("json", {})))
resp = MagicMock()
resp.json.return_value = {"status": "waived", "error": None}
return resp
client.post.side_effect = _post
return client
finding_id = "11111111-1111-1111-1111-111111111111"
with patch("httpx.Client", side_effect=_client_factory):
result = flow_module_auditor.waive_finding(finding_id, "cosmetic nit")
assert result["status"] == "waived"
orch_calls = [(u, b) for u, b in captured if "test-orchestrator" in u]
assert len(orch_calls) == 1
url, body = orch_calls[0]
assert url.endswith("/api/v1/flow/auditor/waive_finding"), (
f"waive_finding must POST to /auditor/waive_finding, got {url}"
)
assert body == {"finding_id": finding_id, "note": "cosmetic nit"}
@@ -289,3 +289,55 @@ async def test_mark_waived_requires_note(db_session: AsyncSession) -> None:
async def test_mark_waived_unknown_id_returns_false(db_session: AsyncSession) -> None:
repo = ReviewFindingsRepository(db_session)
assert await repo.mark_waived(uuid4(), "note") is False
@pytest.mark.asyncio
async def test_list_open_findings_cross_task_blocking_first(
db_session: AsyncSession,
) -> None:
"""list_open_findings returns OPEN rows across tasks, blocker first."""
agent_id = await _seed_agent(db_session)
task_a = await _seed_task(db_session, agent_id)
task_b = await _seed_task(db_session, agent_id)
repo = ReviewFindingsRepository(db_session)
await repo.insert_many(
task_id=task_a,
origin="qa",
round=1,
author_slug="be-qa",
findings=[_finding(severity=Severity.MINOR)],
)
await repo.insert_many(
task_id=task_b,
origin="qa",
round=1,
author_slug="be-qa",
findings=[_finding(severity=Severity.BLOCKER)],
)
# Waive the minor one so only the blocker is OPEN.
open_rows = await repo.list_for_task(task_a, status=STATUS_OPEN)
await repo.mark_waived(UUID(str(open_rows[0].id)), "waived in test")
result = await repo.list_open_findings(limit=20)
assert len(result) == 1
assert result[0].severity == Severity.BLOCKER.value
assert str(result[0].task_id) == str(task_b)
@pytest.mark.asyncio
async def test_list_open_findings_excludes_non_open(
db_session: AsyncSession,
) -> None:
agent_id = await _seed_agent(db_session)
task_id = await _seed_task(db_session, agent_id)
repo = ReviewFindingsRepository(db_session)
rows = await repo.insert_many(
task_id=task_id,
origin="qa",
round=1,
author_slug="be-qa",
findings=[_finding(severity=Severity.NIT)],
)
await repo.mark_waived(UUID(str(rows[0].id)), "nit, skip")
assert await repo.list_open_findings(limit=20) == []