Files
PaperClipAI/ui/src/components/IssueRow.tsx
T
cb0009b097 fix: preserve recovery retries across restarts (#11817)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - The control plane must keep each active issue on a clear execution
or recovery path.
> - A missing issue disposition can require more than one bounded repair
attempt.
> - A server restart could lose that repair path or move source
ownership to the recovery owner.
> - A parked or expired retry could also make the user interface show a
false healthy state.
> - Concurrent recovery loops must not schedule the same repair attempt
twice.
> - This pull request keeps retry state durable, makes scheduling
atomic, and keeps source ownership stable.
> - The benefit is that recovery continues after a restart and operators
see the correct state.

## Linked Issues or Issue Description

**What happened?**

A run that ended without a valid issue disposition could lose its repair
path after a server restart. Manager recovery could also change the
source owner. In addition, a parked or expired retry could make the
issue look healthy when no active work existed. Concurrent
reconciliation could also schedule the same repair attempt twice.

**Expected behavior**

Paperclip must keep bounded source and manager repair attempts across
restarts. Recovery ownership must stay separate from source issue
ownership. The server and user interface must report only a live retry
as active work. Each repair attempt must be scheduled at most once per
company.

**Steps to reproduce**

1. Start an agent run on an issue.
2. End the run without a valid issue disposition.
3. Let the first repair attempt schedule a retry.
4. Restart the server, let the retry time pass without a live run, or
start two reconciliation loops together.
5. Observe that the repair path can stop, the issue can show a false
healthy state, or duplicate retries can be created.

**Paperclip version or commit**

The problem existed on `master` before candidate head
`d8e620fe86bade7df18decac332007f5821ae04f`.

**Deployment mode**

The problem affects self-hosted servers and local builds that use
automatic recovery.

## What Changed

- Persist bounded source-owner and manager repair lineages with stable
fingerprints and retry limits.
- Resume incomplete disposition repairs after a server restart.
- Keep recovery ownership separate from source issue ownership and
enforce source mutation authority.
- Project live retry evidence into issue and blocker summaries.
- Show recovery owner, return owner, attempt count, and retry state in
the board user interface.
- Treat expired or parked retries as attention states unless a queued or
running attempt exists.
- Atomically deduplicate disposition-repair wake requests with a
company-scoped partial unique index.
- Reuse the winning run when concurrent reconciliation loses the
uniqueness race, without duplicate scheduling activity.
- Honor disabled on-demand wake policy before recovery scheduling and
again before delayed retry promotion.
- Keep the new index migration safe for lagging seeded databases that
already contain the index.
- Add server and user interface tests for recovery, restart, ownership,
retry, concurrency, and blocker states.
- Update the implementation and execution semantics documents.

## Verification

- Focused server recovery and ownership suites: 282 tests passed on the
repaired base candidate.
- Focused user interface recovery suites: 128 tests passed on the
repaired base candidate.
- Atomic-deduplication schema and recovery suites: 111 tests passed on
the first Greptile repair.
- Recovery and scheduled-retry wake-policy suites: 126 tests passed at
`d8e620fe86bade7df18decac332007f5821ae04f`.
- The exact lagging-source migration-order test passed after the index
migration became idempotent: 1 test passed and 62 unrelated tests were
skipped.
- `@paperclipai/db` and `@paperclipai/server` typechecks passed at the
current head.
- Migration generation and migration safety checks passed for migration
`0226_tan_colossus.sql`.
- `pnpm check:token-gates` passed on the repaired base candidate.
- `pnpm -r typecheck` passed on the repaired base candidate.
- `pnpm build` passed on the repaired base candidate.
- `pnpm test:run` passed 4,540 tests on the repaired base candidate.
Four fixed-port cases met listeners that already existed on the host.
- The two unchanged fixed-port files passed in an isolated network
namespace: 129 tests passed and 27 tests were skipped.
- Independent Security and QA reviews approved
`63c0423aab54c66f2293a20b0fb3f3b013ee3ba8`; exact-head re-review is
required after automated checks settle on
`d8e620fe86bade7df18decac332007f5821ae04f`.

## Risks

- Recovery orchestration affects issue liveness and ownership. The new
paths use bounded attempts, stable fingerprints, row locks, authority
checks, and database uniqueness.
- A conservative attention state can show more warnings when a scheduled
retry has no queued or running attempt. It does not hide stopped work.
- Migration `0226_tan_colossus.sql` creates a partial unique index on a
known-large table. Migrations run transactionally, so `CONCURRENTLY` is
unavailable. The matching disposition-repair key namespace is introduced
by this release, so deployed databases have no matching rows before the
index is added.

> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.

## Model Used

- OpenAI Codex from the GPT-5 model family used agentic reasoning, tool
use, and code execution. The runtime did not expose the exact model ID
or context window.
- Anthropic Claude Opus 5 used a 1M context window, tool use, and code
execution for part of the user interface repair, as recorded in the
commit history.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

---------

Co-authored-by: Paperclip <noreply@paperclip.ing>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-20 17:09:42 -05:00

393 lines
16 KiB
TypeScript

import type { ReactNode } from "react";
import type { ExternalObjectSummary, Issue, IssueRecoveryAction } from "@paperclipai/shared";
import { Link } from "@/lib/router";
import { Archive, Eye, Flag } from "lucide-react";
import {
createIssueDetailPath,
rememberIssueDetailLocationState,
withIssueDetailHeaderSeed,
} from "../lib/issueDetailBreadcrumb";
import { cn } from "../lib/utils";
import {
deriveActiveRecoveryDisplayState,
RECOVERY_CHIP_DEFAULT_TONE,
recoveryChipLabel,
} from "../lib/recovery-display";
import {
formatRecoveryLineageSummary,
readRecoveryRetryLineage,
type RecoveryLivenessContext,
} from "../lib/recovery-lineage";
import { StatusIcon } from "./StatusIcon";
import { productivityReviewTriggerLabel } from "./ProductivityReviewBadge";
import { hasAssignedBacklogBlocker } from "../lib/issue-blockers";
import { ExternalObjectStatusSummary } from "./ExternalObjectStatusSummary";
import { Badge } from "@/components/ui/badge";
type UnreadState = "hidden" | "visible" | "fading";
interface IssueRowProps {
issue: Issue;
issueLinkState?: unknown;
selected?: boolean;
mobileLeading?: ReactNode;
desktopMetaLeading?: ReactNode;
desktopLeadingSpacer?: boolean;
mobileMeta?: ReactNode;
desktopTrailing?: ReactNode;
/**
* Optional pre-fetched external-object summary. Renders a compact severity
* marker before the rest of `desktopTrailing` on desktop only.
*/
externalObjectSummary?: ExternalObjectSummary | null;
trailingMeta?: ReactNode;
titleSuffix?: ReactNode;
titleClassName?: string;
checklistStepNumber?: number | string | null;
checklistCurrentStep?: boolean;
checklistDependencyChips?: ReactNode;
checklistRowId?: string;
unreadState?: UnreadState | null;
onMarkRead?: () => void;
onArchive?: () => void;
archiveDisabled?: boolean;
className?: string;
/** Pointer entered the row (used by list keyboard nav to track hover). */
onMouseEnter?: () => void;
/** Ancestor levels; renders that many vertical tree-guide slots (desktop). */
treeGuides?: number;
/**
* This row has its own collapse chevron sitting in the innermost guide
* column (a nested parent). Breaks the guide line there so the chevron is
* not crossed out by it.
*/
chevronInGuide?: boolean;
/** Opt in to a bottom divider on this row (default off; used by views that intentionally keep separators). */
showDivider?: boolean;
}
export function InboxArchiveButton({
onArchive,
disabled,
}: {
onArchive: () => void;
disabled?: boolean;
}) {
return (
<button
type="button"
data-slot="icon-button"
onClick={(event) => {
event.preventDefault();
event.stopPropagation();
onArchive();
}}
onKeyDown={(event) => {
if (event.key !== "Enter" && event.key !== " ") return;
event.preventDefault();
event.stopPropagation();
onArchive();
}}
disabled={disabled}
className="inline-flex shrink-0 items-center gap-1.5 rounded-md px-2 py-1 text-xs font-medium text-muted-foreground opacity-0 transition-opacity hover:bg-accent hover:text-foreground group-hover:opacity-100 focus-visible:opacity-100 disabled:pointer-events-none disabled:opacity-30"
aria-label="Archive"
>
<Archive className="h-3.5 w-3.5" />
Archive
</button>
);
}
export function IssueRow({
issue,
issueLinkState,
selected = false,
mobileLeading,
desktopMetaLeading,
desktopLeadingSpacer = false,
mobileMeta,
desktopTrailing,
externalObjectSummary,
trailingMeta,
titleSuffix,
titleClassName,
checklistStepNumber = null,
checklistCurrentStep = false,
checklistDependencyChips,
checklistRowId,
unreadState = null,
onMarkRead,
onArchive,
archiveDisabled,
className,
onMouseEnter,
treeGuides = 0,
chevronInGuide = false,
showDivider = false,
}: IssueRowProps) {
const issuePathId = issue.identifier ?? issue.id;
const identifier = issue.identifier ?? issue.id.slice(0, 8);
// A row participates in the unread system whenever `unreadState` is supplied
// (inbox rows). It then reserves a fixed leading dot slot on all rows — read
// and unread alike — so the mark-read dot sits in the far-left gutter without
// shifting content, matching the sibling non-issue inbox rows.
const showUnreadSlot = unreadState != null;
const showUnreadDot = unreadState === "visible" || unreadState === "fading";
const unreadDotButton = (
<button
type="button"
data-slot="icon-button"
onClick={(event) => {
event.preventDefault();
event.stopPropagation();
onMarkRead?.();
}}
onKeyDown={(event) => {
if (event.key === "Enter" || event.key === " ") {
event.preventDefault();
event.stopPropagation();
onMarkRead?.();
}
}}
className={cn(
"inline-flex h-4 w-4 items-center justify-center rounded-full transition-colors",
selected ? "hover:bg-muted/80" : "hover:bg-blue-500/20",
)}
aria-label="Mark as read"
>
<span
className={cn(
"block h-2 w-2 rounded-full transition-opacity duration-300",
selected ? "bg-muted-foreground/70" : "bg-blue-600 dark:bg-blue-400",
unreadState === "fading" ? "opacity-0" : "opacity-100",
)}
/>
</button>
);
const selectedStatusClass = selected ? "!text-muted-foreground !border-muted-foreground" : undefined;
const detailState = withIssueDetailHeaderSeed(issueLinkState, issue);
const productivityReview = issue.productivityReview ?? null;
const productivityReviewIndicator = productivityReview ? (
<span
className={cn(
"inline-flex h-4 w-4 shrink-0 items-center justify-center rounded-full border border-amber-500/40 bg-amber-500/10 text-amber-600 dark:text-amber-300",
selected ? "border-muted-foreground text-muted-foreground" : null,
)}
title={`Productivity review: ${productivityReviewTriggerLabel(productivityReview.trigger)}`}
aria-label="Productivity review open"
>
<Eye className="h-2.5 w-2.5" aria-hidden />
</span>
) : null;
const hasChecklistStep = checklistStepNumber !== null;
const checklistStep = hasChecklistStep ? (
<span className="shrink-0 font-mono text-xs text-muted-foreground" aria-hidden="true">
{checklistStepNumber}.
</span>
) : null;
const recoveryAction = issue.activeRecoveryAction ?? null;
// The row already carries the issue's own scheduled retry, so the chip can tell a retry the
// scheduler is actually running from one whose due time simply passed.
const recoveryIndicator = recoveryAction
? renderRecoveryChip(recoveryAction, selected, { scheduledRetry: issue.scheduledRetry ?? null })
: null;
const parkedBlockerIndicator = hasAssignedBacklogBlocker(issue.blockedBy) ? (
<Badge variant="outline"
data-testid="issue-row-parked-blocker"
className="[&>svg]:size-2.5 ml-1.5 gap-0.5 border-amber-500/60 bg-amber-500/15 text-(length:--text-nano) text-amber-700 dark:text-amber-300"
title="Blocked by parked work — at least one assigned blocker is in backlog and will not wake its assignee."
>
<Flag className="h-2.5 w-2.5" aria-hidden />
Blocked by parked work
</Badge>
) : null;
return (
<div
onMouseEnter={onMouseEnter}
className={cn(
// No color transition on the row band: hover/selection must snap
// instantly. A fade (transition-colors) leaves a trail of fading bands
// when scrubbing the mouse fast across the list.
"group relative flex items-start gap-2 rounded-lg py-2.5 pr-3 text-sm no-underline text-inherit sm:items-center sm:py-2 sm:pl-1",
showUnreadSlot ? "pl-4" : "pl-2",
"[&_button]:relative [&_button]:z-10",
// Divider + hover/selected/checklist wash live on the ROOT row band so
// the tint paints BEHIND the content and `last:border-b-0` matches the
// real last row. Keeping these on the overlay Link (PR #10526) made the
// last row keep its border and the hover wash paint over the text.
showDivider && "border-b border-border last:border-b-0",
selected ? "hover:bg-transparent" : "hover:bg-accent/50",
checklistCurrentStep ? "bg-primary/5" : null,
className,
)}
>
<Link
to={createIssueDetailPath(issuePathId)}
state={detailState}
disableIssueQuicklook
issuePrefetch={issue}
data-inbox-issue-link
id={checklistRowId}
aria-current={checklistCurrentStep ? "step" : undefined}
onClickCapture={() => rememberIssueDetailLocationState(issuePathId, detailState)}
className={cn(
// Overlay Link keeps ONLY positioning + focus ring so header controls
// stay clickable above it; visual washes belong on the root above.
"absolute inset-0 rounded-lg no-underline text-inherit focus-visible:z-10 focus-visible:outline-none focus-visible:ring-(length:--rad-3) focus-visible:ring-ring",
)}
>
<span className="sr-only">Open {identifier}: {issue.title}</span>
</Link>
<span className="flex shrink-0 items-center gap-1 pt-px sm:hidden">
{mobileLeading ?? <StatusIcon status={issue.status} blockerAttention={issue.blockerAttention} size="md" className={selectedStatusClass} />}
{productivityReviewIndicator}
{parkedBlockerIndicator}
{recoveryIndicator}
</span>
<span className="flex min-w-0 flex-1 flex-col gap-1 sm:contents">
<span className={cn("line-clamp-2 text-sm sm:order-2 sm:min-w-0 sm:flex-1 sm:truncate sm:line-clamp-none", titleClassName)}>
{issue.title}{titleSuffix}
</span>
{checklistDependencyChips ? (
<span className="flex flex-wrap gap-1 sm:order-3 sm:ml-(--sz-calc-13)">
{checklistDependencyChips}
</span>
) : null}
<span className="flex items-center gap-2 self-stretch sm:order-1 sm:shrink-0">
{showUnreadSlot ? (
// Reserved leftmost dot gutter (desktop). Present on read and unread
// rows so the mark-read dot lives to the LEFT of any leading control
// (a parent's collapse caret, a tree guide) without indenting the row
// relative to its siblings, and aligns with the non-issue inbox rows
// that reserve the same w-4 slot.
<span
data-testid="issue-row-unread-slot"
className="hidden h-4 w-4 shrink-0 items-center justify-center self-center sm:inline-flex"
>
{showUnreadDot ? unreadDotButton : null}
</span>
) : null}
{treeGuides > 0
? Array.from({ length: treeGuides }, (_, level) => {
// The innermost guide lands on THIS row's own chevron column; if
// the row has a chevron, break the line around it so it isn't
// crossed out.
const gapForChevron = chevronInGuide && level === treeGuides - 1;
return (
// Tree guide: occupies the same flex slot as the parent's
// chevron column so the line lands under the parent's status
// column; stretched past the row padding so consecutive rows
// read as one continuous line.
<span key={`guide-${level}`} aria-hidden="true" className="relative hidden w-4 shrink-0 self-stretch sm:block">
{/* The connector drops from under the ancestor's STATUS icon,
not its chevron: the status column sits one level (w-4 slot
+ gap-2 = 2rem) right of this guide slot's left edge.
bg-background underlay: dark-mode --border is translucent,
so overlapping row segments would stack brighter without
an opaque base. */}
<span className="absolute -inset-y-3 left-8 w-px bg-background">
{gapForChevron ? (
// Two border segments centering a 14px (h-3.5) transparent
// gap for the row's own chevron.
<span className="absolute inset-0 flex flex-col">
<span className="flex-1 bg-border" />
<span className="h-3.5 shrink-0" />
<span className="flex-1 bg-border" />
</span>
) : (
<span className="absolute inset-0 bg-border" />
)}
</span>
</span>
);
})
: null}
{desktopLeadingSpacer ? (
<span className="hidden w-3.5 shrink-0 sm:block" />
) : null}
{desktopMetaLeading ?? (
<>
<span className="hidden shrink-0 items-center gap-1 sm:inline-flex">
<StatusIcon status={issue.status} blockerAttention={issue.blockerAttention} size="md" className={selectedStatusClass} />
{productivityReviewIndicator}
</span>
{checklistStep}
<span className="shrink-0 font-mono text-xs text-muted-foreground">
{identifier}
</span>
{parkedBlockerIndicator}
{recoveryIndicator}
</>
)}
{mobileMeta ? (
<>
<span className="text-xs text-muted-foreground sm:hidden" aria-hidden="true">
&middot;
</span>
<span className="text-xs text-muted-foreground sm:hidden">{mobileMeta}</span>
</>
) : null}
</span>
</span>
{(onArchive || desktopTrailing || trailingMeta || externalObjectSummary) ? (
<span className="ml-auto hidden shrink-0 items-center gap-2 sm:order-3 sm:flex sm:gap-3">
{onArchive ? (
<InboxArchiveButton onArchive={onArchive} disabled={archiveDisabled} />
) : null}
{externalObjectSummary ? (
<ExternalObjectStatusSummary summary={externalObjectSummary} compact />
) : null}
{desktopTrailing}
{trailingMeta ? (
<span className="text-xs text-muted-foreground">{trailingMeta}</span>
) : null}
</span>
) : null}
{showUnreadDot ? (
// Inbox rows reserve a mobile gutter on both read and unread rows. The
// full control stays inside overflow-clipping row containers while its
// absolute position avoids shifting or covering the leading control.
<span className="absolute left-0 top-1/2 inline-flex h-4 w-4 -translate-y-1/2 items-center justify-center sm:hidden">
{unreadDotButton}
</span>
) : null}
</div>
);
}
function renderRecoveryChip(
action: IssueRecoveryAction,
selected: boolean,
liveness: RecoveryLivenessContext,
): ReactNode {
const state = deriveActiveRecoveryDisplayState(action, liveness);
if (!state) return null;
const tone = RECOVERY_CHIP_DEFAULT_TONE[state];
const Icon = tone.icon;
const lineage = readRecoveryRetryLineage(action, liveness);
const label = recoveryChipLabel(state, action.kind, lineage);
const detail = lineage ? formatRecoveryLineageSummary(lineage) : null;
return (
<Badge variant="outline"
data-testid="issue-row-recovery-indicator"
data-recovery-state={state}
data-recovery-kind={action.kind}
data-recovery-lane={lineage?.lane}
role="status"
aria-label={detail ? `${label} — ${detail}` : label}
className={cn(
"ml-1.5 gap-0.5 text-(length:--text-nano)",
tone.className,
selected ? "!border-muted-foreground !text-muted-foreground" : null,
)}
title={detail
? `${label} — ${detail}. Open the source task to act.`
: `${label} — open the source task to act.`}
>
<Icon className="h-2.5 w-2.5" aria-hidden />
{label}
</Badge>
);
}