Collectors apply RULE_KEYS (or index stems) into fsm_tables before normalize; parsers prefer FSM or hand-parse.
Co-authored-by: Cursor <cursoragent@cursor.com>
Introduce ParseProfile-based business state tasks that reuse LLDP parsing for snapshot batches (separate from topology Fabric writes), with scheduler, export, and a Phase1 UI under network tasks.
Co-authored-by: Cursor <cursoragent@cursor.com>
Refresh left-nav/main chrome, unify list status chips across network pages, and merge UME token with alarm subscription into one top card.
Co-authored-by: Cursor <cursoragent@cursor.com>
Key alerts now deliver only via the DSH hub; Modal Body no longer flex-shrinks the rules rows to empty.
Co-authored-by: Cursor <cursoragent@cursor.com>
Pull top-5 matches from UME inventory, expand the list in-flow under the input, and auto-scroll the modal to keep results visible.
Co-authored-by: Cursor <cursoragent@cursor.com>
Drop fabric reconcile as redundant with topology, open the traffic wall from the sched card, and surface scheduler tick ages in metrics types.
Co-authored-by: Cursor <cursoragent@cursor.com>
Checkpoint before workbench facade redesign: list defaults, API key quota, toast portal, topology canvas editor toolbar with More menu, and trusted-proxy client IP for sessions.
Co-authored-by: Cursor <cursoragent@cursor.com>
Track per-connection metadata on the hub and surface it in the UME key-alert UI so operators can see which Netx Ops clients are subscribed.
Co-authored-by: Cursor <cursoragent@cursor.com>
Canonical playbooks live under skills/; netx-mcp 0.3.0 exposes queryNmsAlarms et al. with nms_ne_id aliases.
Co-authored-by: Cursor <cursoragent@cursor.com>
netx hosts /v1/integrations/dsh-alarm/ws; matched key alerts fan out to dialed-in DSH subscribers while the optional oclaw bridge remains.
Co-authored-by: Cursor <cursoragent@cursor.com>
Drain leftover prompts before Netmiko send_command, and make stop/start tolerate orphan workers and locked log files.
Co-authored-by: Cursor <cursoragent@cursor.com>
Fix bare Enter re-audits, Tab expansion lag, and history mid-line edits by preferring the xterm/device row at Enter and rejecting empty-prompt snapshots.
Co-authored-by: Cursor <cursoragent@cursor.com>
Audit every completed stdin line per flush, keep a list of audit_line hints from the router, restore prompt enrichment and stdout fallback, and record actual bytes when hints disagree with what was sent.
Co-authored-by: Cursor <cursoragent@cursor.com>
Only audit interactive commands from frontend audit_line with a CLI prompt prefix; skip prompt-sync Enter and device error echoes; join wrapped xterm rows for long commands.
Co-authored-by: Cursor <cursoragent@cursor.com>
xterm and keydown can both emit carriage return for one keypress; skip duplicate Enter frames on the client and collapse identical audits within 450ms on the server.
Co-authored-by: Cursor <cursoragent@cursor.com>
Do not reuse the previous command when the operator presses Enter on a blank prompt; validate prompt-only rows on both frontend and backend.
Co-authored-by: Cursor <cursoragent@cursor.com>
PTY stdout/stdin carry backspace and edit noise during typing; record only the final row the operator sees when pressing Enter.
Co-authored-by: Cursor <cursoragent@cursor.com>
ZTE tab redraws the input line via carriage return; parse the last prompt line from PTY stdout on Enter instead of stdin keystrokes. Also remove remaining ?? in AuditPage.
Co-authored-by: Cursor <cursoragent@cursor.com>
Remove duplicate serializeTerminal fragment that broke WebTerminal compilation; rewrite ne.exec_batch count without mixing ?? and ||.
Co-authored-by: Cursor <cursoragent@cursor.com>
On Enter, send audit_line from the terminal row so tab completion and edits match what the operator executed; keep the device prompt prefix in audit logs.
Co-authored-by: Cursor <cursoragent@cursor.com>
