All checks were successful
G12 Leak-Guard / leak-guard (pull_request) Successful in 6s
פאנל ב-/operations שמראה אילו סוכני Paperclip עובדים כעת (רצים+בתור), הפלט החי
שלהם, ושליטה מבוקרת: עצירת ריצה, איפוס session. סוגר את הנקודה-העיוורת שבה drain
מונע-סוכן (למשל ריקון תור הלכות ע"י ה-CEO heartbeat) עוקף את בקרת /operations
שמכירה רק שירותי pm2, והפלט הגולמי נגיש רק ב-Paperclip UI.
מקור-נתונים: Paperclip heartbeat-runs API (אומת חי):
GET /api/companies/{cid}/live-runs — רצים+בתור (agentName/status/issue/outputSilence)
GET /api/heartbeat-runs/{id}/log — NDJSON של פלט הסוכן
GET /api/heartbeat-runs/{id}/events — timeline
POST /api/heartbeat-runs/{id}/cancel — עצירה מבוקרת (לא kill — מכבד watchdog+checkpoint)
POST /api/agents/{id}/runtime-state/reset-session
ארכיטקטורה (G12/INV-PORT1): כל המגע החדש עם Paperclip דרך השער בלבד —
web/paperclip_client.py (shell) → re-export ב-web/agent_platform_port.py →
web/app.py צורך מהשער. leak_guard.py עובר (seam שלם). אסור kill ישיר על
process_pid (עוקף את השער).
Backend:
- paperclip_client: list_live_runs / get_run_log / get_run_events / cancel_run / reset_agent_session
- agent_platform_port: re-export pc_list_live_runs / pc_get_run_log / pc_get_run_events / pc_cancel_run / pc_reset_agent_session
- app.py: GET /api/operations/agents (אגרגציה CMP+CMPA, עמיד לכשל-חברה),
GET .../runs/{id}/log, GET .../runs/{id}/events, POST .../runs/{id}/cancel,
POST .../agents/{id}/reset-session
Frontend: פאנל "סוכנים פעילים" ב-/operations (polling 4s) + dialog ללוג חי
(פרסור NDJSON→טקסט קריא) + כפתורי עצור/אפס. הוספת hooks ל-operations.ts.
בטיחות: cancel על דריינר הלכות בטוח — חילוץ checkpointed per-chunk + resumable
+ self-heal לשורות processing.
Invariants: מקיים G12/INV-PORT1 (שער-הפלטפורמה). נוגע X6 (UI↔API).
api:types יורץ אחרי deploy (openapi.json חי).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
191 lines
5.9 KiB
TypeScript
191 lines
5.9 KiB
TypeScript
import { useMutation, useQuery, useQueryClient } from "@tanstack/react-query";
|
|
import { toast } from "sonner";
|
|
import { apiRequest } from "./client";
|
|
|
|
export type OpsService = {
|
|
name: string;
|
|
status: string;
|
|
restarts: number;
|
|
uptime_ms: number;
|
|
cpu: number;
|
|
memory_bytes: number;
|
|
cron: string;
|
|
autorestart: boolean;
|
|
disabled?: boolean; // cron drain switched off via the dashboard
|
|
};
|
|
|
|
export type CourtFetchJob = {
|
|
case_number_norm: string;
|
|
citation_raw: string;
|
|
tier: string;
|
|
status: string;
|
|
error: string;
|
|
updated_at: string;
|
|
};
|
|
|
|
export type IngestedRow = {
|
|
case_number: string;
|
|
court: string;
|
|
source_url: string;
|
|
created_at: string;
|
|
};
|
|
|
|
/** The uniform per-pipeline shape every background drain reports. */
|
|
export type PipelineStats = {
|
|
pending: number; // backlog: rows not yet processed (status default)
|
|
processing: number; // being worked right now
|
|
done: number; // completed
|
|
failed: number; // terminal failures (court_fetch folds in 'manual')
|
|
queued: number; // explicitly enqueued for the next drain run
|
|
running_now: string[]; // human labels of the items currently processing
|
|
by_status: Record<string, number>; // raw counts, for the curious
|
|
};
|
|
|
|
export type OperationsSnapshot = {
|
|
services: OpsService[];
|
|
services_error: string | null;
|
|
pipelines: {
|
|
court_fetch: PipelineStats & { recent: CourtFetchJob[] };
|
|
metadata_extraction: PipelineStats;
|
|
halacha_extraction: PipelineStats;
|
|
digests: PipelineStats & { total: number; linked: number };
|
|
halacha_review: { by_status: Record<string, number> };
|
|
missing_precedents: { by_status: Record<string, number> };
|
|
ingested_recent: IngestedRow[];
|
|
};
|
|
};
|
|
|
|
export function useOperations() {
|
|
return useQuery({
|
|
queryKey: ["operations"],
|
|
queryFn: ({ signal }) =>
|
|
apiRequest<OperationsSnapshot>("/api/operations", { signal }),
|
|
refetchInterval: 5000, // live view of background work
|
|
staleTime: 3000,
|
|
});
|
|
}
|
|
|
|
export type ServiceAction = "restart" | "stop" | "start" | "run-now";
|
|
|
|
/** Control a background service (daemon restart/stop/start, or run a drain now). */
|
|
export function useServiceAction() {
|
|
const qc = useQueryClient();
|
|
return useMutation({
|
|
mutationFn: ({ name, action }: { name: string; action: ServiceAction }) =>
|
|
apiRequest(`/api/operations/services/${name}/${action}`, { method: "POST" }),
|
|
onSuccess: (_d, { action }) => {
|
|
const labels: Record<ServiceAction, string> = {
|
|
"run-now": "הופעל עכשיו",
|
|
restart: "הופעל מחדש",
|
|
stop: "נעצר",
|
|
start: "הופעל",
|
|
};
|
|
toast.success(labels[action]);
|
|
qc.invalidateQueries({ queryKey: ["operations"] });
|
|
},
|
|
onError: (e) => toast.error(`הפעולה נכשלה: ${String(e)}`),
|
|
});
|
|
}
|
|
|
|
/** Switch a cron drain on/off (its "startup type"). */
|
|
export function useDrainToggle() {
|
|
const qc = useQueryClient();
|
|
return useMutation({
|
|
mutationFn: ({ name, disabled }: { name: string; disabled: boolean }) =>
|
|
apiRequest(`/api/operations/drains/${name}/disabled`, {
|
|
method: "POST",
|
|
body: { disabled },
|
|
}),
|
|
onSuccess: (_d, { disabled }) => {
|
|
toast.success(disabled ? "התזמון כובה" : "התזמון הופעל");
|
|
qc.invalidateQueries({ queryKey: ["operations"] });
|
|
},
|
|
onError: (e) => toast.error(`העדכון נכשל: ${String(e)}`),
|
|
});
|
|
}
|
|
|
|
// ── Live agents — which agent is working now + its output + controls ───────
|
|
|
|
export type AgentRun = {
|
|
run_id: string;
|
|
agent_id: string;
|
|
agent_name: string;
|
|
company_id: string;
|
|
company_label: string;
|
|
status: string; // running | queued | ...
|
|
invocation_source: string;
|
|
trigger_detail: string;
|
|
issue_id: string | null;
|
|
adapter_type: string;
|
|
started_at: string | null;
|
|
created_at: string | null;
|
|
last_output_at: string | null;
|
|
continuation_attempt: number;
|
|
silence_level: string; // "" | ok | suspicion | critical
|
|
silence_age_ms: number;
|
|
};
|
|
|
|
export type AgentRunsResponse = {
|
|
runs: AgentRun[];
|
|
running: number;
|
|
queued: number;
|
|
errors: string[];
|
|
};
|
|
|
|
export type RunLog = {
|
|
runId: string;
|
|
store: string;
|
|
logRef: string;
|
|
content: string; // NDJSON stream the adapter captured
|
|
};
|
|
|
|
/** Queued + running heartbeat runs across all companies. */
|
|
export function useAgentRuns() {
|
|
return useQuery({
|
|
queryKey: ["operations", "agents"],
|
|
queryFn: ({ signal }) =>
|
|
apiRequest<AgentRunsResponse>("/api/operations/agents", { signal }),
|
|
refetchInterval: 4000, // live view of who's working now
|
|
staleTime: 2000,
|
|
});
|
|
}
|
|
|
|
/** Full output log of one run — fetched on demand (drawer open). */
|
|
export function useRunLog(runId: string | null) {
|
|
return useQuery({
|
|
queryKey: ["operations", "agents", "log", runId],
|
|
queryFn: ({ signal }) =>
|
|
apiRequest<RunLog>(`/api/operations/agents/runs/${runId}/log`, { signal }),
|
|
enabled: !!runId,
|
|
refetchInterval: runId ? 4000 : false, // live tail while open
|
|
});
|
|
}
|
|
|
|
/** Gracefully cancel a queued/running run (not a raw kill). */
|
|
export function useCancelRun() {
|
|
const qc = useQueryClient();
|
|
return useMutation({
|
|
mutationFn: (runId: string) =>
|
|
apiRequest(`/api/operations/agents/runs/${runId}/cancel`, { method: "POST" }),
|
|
onSuccess: () => {
|
|
toast.success("בקשת עצירה נשלחה");
|
|
qc.invalidateQueries({ queryKey: ["operations", "agents"] });
|
|
},
|
|
onError: (e) => toast.error(`העצירה נכשלה: ${String(e)}`),
|
|
});
|
|
}
|
|
|
|
/** Reset a wedged agent session so its next wakeup starts clean. */
|
|
export function useResetAgentSession() {
|
|
const qc = useQueryClient();
|
|
return useMutation({
|
|
mutationFn: (agentId: string) =>
|
|
apiRequest(`/api/operations/agents/${agentId}/reset-session`, { method: "POST" }),
|
|
onSuccess: () => {
|
|
toast.success("ה-session אופס");
|
|
qc.invalidateQueries({ queryKey: ["operations", "agents"] });
|
|
},
|
|
onError: (e) => toast.error(`האיפוס נכשל: ${String(e)}`),
|
|
});
|
|
}
|