pbakaus_impeccable/site/data/live-harnesses.ts

export const harnessPaths = [
  {
    name: 'Dedicated app-server worker',
    harness: 'Codex · isolated persistent thread',
    pickup: '1.47 ms turn wake',
    tradeoff: 'A separate Live-owned Sol thread generates while the control supervisor keeps polling. Full context protects design quality; cold readiness measured 657 ms and happens during Live initialization.',
    status: 'Codex default · gated',
    statusTone: 'ready',
    source: 'https://learn.chatgpt.com/docs/app-server',
  },
  {
    name: 'Foreground terminal supervisor',
    harness: 'Codex desktop · unified exec session',
    pickup: 'Concurrent shell work proven',
    tradeoff: 'A yielded foreground process keeps running while other commands execute, but new output still needs an explicit session read; it does not wake a finished model turn by itself.',
    status: 'Useful within active turn',
    statusTone: 'measure',
    source: 'https://learn.chatgpt.com/docs/app-server#api-overview',
  },
  {
    name: 'Native generation subagent',
    harness: 'Codex desktop · portable fallback',
    pickup: 'Main turn must dispatch it',
    tradeoff: 'Low-effort compact handoff is safe and fenced, but it cannot start after the desktop task has gone idle.',
    status: 'Current fallback',
    statusTone: 'prototype',
    source: 'https://learn.chatgpt.com/docs/agent-configuration/subagents',
  },
  {
    name: 'Background variant producer',
    harness: 'Claude Code subagent',
    pickup: 'Supported; measure per version',
    tradeoff: 'Model, effort, background execution, and resume are configurable. Preserve the current main-thread path as fallback.',
    status: 'Cross-harness control',
    statusTone: 'control',
    source: 'https://code.claude.com/docs/en/sub-agents',
  },
];

export const liveExperiments = [
  {
    rank: 1,
    title: 'Run a dedicated Codex worker — shipped',
    evidence: 'Matched full tasks reached usable output in 34.9 s warm app-server versus 64.6 s direct, with both paths passing 4/4. Live now returns from prewarm in 87 ms while the worker finishes initializing alongside page startup.',
    move: 'Keep one Live-owned app-server connection and isolated Sol/medium thread. Attach the skill, sub-command, project context, source neighborhood, and annotation image.',
    expected: 'Removes main-turn availability from generation without trading away brand fidelity.',
    confidence: 'Production browser + full-task gate',
  },
  {
    rank: 2,
    title: 'Accept any arrived variant — shipped',
    evidence: 'A real Vite/React + Sol run accepted variant 1, interrupted variants 2–3, returned the page in 299 ms, and dispatched the next Go 903 ms after Accept.',
    move: 'Durably fence generation on Accept/Discard, prioritize control events, publish through an epoch + source-hash transaction, and keep a separate poll supervisor active.',
    expected: 'Removes the full-set wait and keeps the page interactive while canceled work unwinds.',
    confidence: 'Production browser',
  },
  {
    rank: 3,
    title: 'Reveal variants progressively — shipped',
    evidence: 'The production run showed variant 1 at 9.87 s and the full set at 31.2 s. Progressive publication removed 21.3 s from the first decision.',
    move: 'Codex publishes a monotonically growing prefix as each variant validates. The browser reveals arrived variants and pending dots immediately; Accept and Discard fence unfinished work.',
    expected: 'Lets review begin as soon as one credible option exists.',
    confidence: 'Production browser',
  },
  {
    rank: 4,
    title: 'Dispatch first, capture second — shipped',
    evidence: 'Plain click-handler → generate fetch is 2.2 ms median; end-to-end model-free latency fell from 916 ms to 414 ms.',
    move: 'Unannotated picks dispatch before off-path shader capture. Annotated picks still upload their visual evidence before dispatch.',
    expected: 'Removes screenshot work from the common critical path.',
    confidence: 'Ten-run browser benchmark',
  },
  {
    rank: 5,
    title: 'Preflight source locally — shipped',
    evidence: 'The production worker resolved and wrapped the selected React source in 52 ms before the model turn.',
    move: 'Attach durable scaffold metadata to the queued event. Every harness reuses it; discovery failures preserve the existing agent-driven fallback.',
    expected: 'Removes one deterministic source-discovery tool round trip from every successful generation.',
    confidence: 'Production browser',
  },
  {
    rank: 6,
    title: 'Generate knobs after pixels — shipped',
    evidence: 'The first progressive write intentionally carries no parameter manifest; tune controls arrive with the complete variant set.',
    move: 'Plan parameter axes with the trio, but defer their manifests and CSS branches until the final delivery edit.',
    expected: 'Keeps parameter output off the first-reviewable critical path without shrinking the final tuning surface.',
    confidence: 'Protocol shipped',
  },
  {
    rank: 7,
    title: 'Acknowledge first, clean up off-path — shipped',
    evidence: 'The production page returned to Pick in 299 ms and dispatched new work while carbonize cleanup remained on its separate foreground control lane.',
    move: 'Keep the durable accept acknowledgement on the foreground path, then hand carbonize cleanup to a source-locked worker while the poll supervisor leases new work.',
    expected: 'Preserves immediate interaction without leaving temporary Live source behind.',
    confidence: 'Production browser + build control',
  },
  {
    rank: 8,
    title: 'Trim the producer contract — selected',
    evidence: 'Progressive compact passed 5/5 for Claude Sonnet 4.6 and GPT-5.5. A repeated warm Sol turn cached 38,656 of 46,973 input tokens (82.3%) and completed in 21.8 s versus 26.0 s cold.',
    move: 'Keep the stable generation contract cacheable and send only the picked-element context plus the current action and identity lock.',
    expected: 'Use provider caching instead of risky context deltas; always send current project and brand evidence.',
    confidence: 'Candidate matrix + token telemetry',
  },
  {
    rank: 9,
    title: 'Parallelize ideas only where reliable — provider-specific',
    evidence: 'Parallel compact passed 5/5 for GPT-5.5 and Gemini 3.1 Flash-Lite, but only 3/5 for Claude Sonnet 4.6.',
    move: 'Use parallel compact for Gemini and as an OpenAI full-set option. Keep progressive compact for Anthropic until its validation reliability improves.',
    expected: 'Lowers full-set latency without shared-file races or pretending one call shape fits every provider.',
    confidence: 'Sixty-run paid matrix',
  },
];

export const currentHarnessProbe = {
  testedAt: '2026-07-13',
  surface: 'Codex desktop unified exec',
  delayedOutputMs: 250,
  processExitMs: null,
  parallelCommandMs: 0.01,
  parallelWorkSucceeded: true,
  backgroundChildSurvivedShellExit: false,
  requiresExplicitSessionRead: true,
  surfacedAutomatically: false,
  result: 'A yielded foreground poller kept producing output while a separate command completed immediately. A traditional shell-backgrounded child did not survive shell exit. Output still required an explicit session read and did not proactively wake a finished turn.',
};