organ-lane-telemetry-cockpit

name	organ-lane-telemetry-cockpit
description	Human-readable cockpit that reads originator-telemetry.jsonl, groups by traceId, and prints a compact per-run timeline with elapsed time, heartbeat count, chars streamed, and status. Use when you need to see what a lane is doing without reading raw JSONL, or when invoking lane-telemetry-cockpit to inspect lane run summaries.
triggers	["organ-lane-telemetry-cockpit","how do I use lane-telemetry-cockpit","lane-telemetry-cockpit usage","lane-telemetry-cockpit guide","Lane Telemetry Cockpit (human-readable view over YURI_LANE_TELEMETRY)"]
generated	true
source_node	lane-telemetry-cockpit
source_file	_SYSTEM/Scripts/lane-telemetry-cockpit.mjs
scope	harness
invocation	workflow

Organ Guide — Lane Telemetry Cockpit (human-readable view over YURI_LANE_TELEMETRY)

Module: _SYSTEM/Scripts/lane-telemetry-cockpit.mjs · Layer: Skills & Orchestration · Invocation: both · CLI: (path) [--json] [--limit N]

Purpose. Human-readable cockpit over the Originator lane-telemetry stream. JSONL is the audit substrate, not the operator experience: this reads originator-telemetry.jsonl, groups by traceId, COLLAPSES the noisy stream/stderr/stdout chunk events into counts, and prints a compact per-run timeline (start → model call → verification → exit) with elapsed, heartbeat count, chars streamed, and status — so the operator sees what a lane is doing without reading raw JSON.

Exports

• readTelemetry(file = TELEMETRY, opts = {})
  • in: optional telemetry file path + opts
  • out: parsed telemetry events (malformed lines tolerated)
• summarizeByTrace(events)
  • in: telemetry events
  • out: per-traceId summary collapsing chunk events into counts (elapsed, heartbeats, chars, status)
• renderCockpit(traces, opts)
  • in: summarized traces + render opts
  • out: the compact human-readable cockpit string

Security boundary

Strictly READ-ONLY — reads the telemetry JSONL and renders it; never mutates telemetry. Input line-count is bounded (limit guard). No protected-path access.

When to use

Inspecting what a dispatched lane run actually did (timeline, model call, verification, exit) without parsing raw originator-telemetry JSON; triaging a slow/failed lane run.

Gotchas

• It collapses stream/stderr/stdout chunk events into COUNTS — individual chunk payloads are not shown (by design, to stay compact).
• Grouping is by traceId; a run without a stable traceId will not group cleanly.
• Read-only over a live-appended file — a snapshot reflects the moment it was read.

Session Notes

2026-06-13

• session: 40m | peak ctx: 0% | compacts: 0
• tools: Bash×824, Read×163, Edit×17, StructuredOutput×16, Write×8, TodoWrite×4, ToolSearch×2, ScheduleWakeup×2, Workflow×1, mcp×1, AskUserQuestion×1
• corrections: none
• errors: none