flynn/scripts/capture-phase0-live-baseline.ts

#!/usr/bin/env node

import { mkdir, writeFile } from 'node:fs/promises';
import { dirname, resolve } from 'node:path';
import { parseArgs } from 'node:util';
import { queryAuditLogs } from '../src/audit/export.js';
import { capturePhase0LiveBaselineEvents } from '../src/audit/phase0LiveBaseline.js';
import {
  renderPhase0BaselineMarkdown,
  summarizePhase0Baseline,
  type AuditSource,
  type Phase0BaselineSummaryOptions,
} from '../src/audit/phase0BaselineSummary.js';

const DEFAULT_EVENT_TYPES = ['run.state', 'run.cancel', 'reaction.match', 'reaction.skip'] as const;

function usage(): string {
  return [
    'Usage: node --import tsx/esm scripts/capture-phase0-live-baseline.ts [options]',
    '',
    'Options:',
    '  --audit <path>                          Source audit log path (default: ~/.local/share/flynn/audit.log)',
    '  --since <ISO-8601|epoch_ms>             Start time filter',
    '  --until <ISO-8601|epoch_ms>             End time filter',
    '  --channel <name[,name...]>              Restrict sample to channels',
    '  --source <gateway|channel[,..]>         Restrict sample to sources',
    '  --exclude-session-substring <text[,..]> Exclude sessions containing any substring (default: probe)',
    '  --raw-identifiers                       Keep raw session/sender/request IDs (default: anonymized)',
    '  --tag <YYYY-MM-DD>                      Output file tag (default: current date UTC)',
    '  --sample-out <path>                     Output JSONL sample path override',
    '  --summary-json-out <path>               Output summary JSON path override',
    '  --summary-md-out <path>                 Output summary markdown path override',
    '  --max-sessions <number>                 Limit session rows in output (default: 20)',
    '  --max-channels <number>                 Limit channel rows in output (default: 20)',
    '  --max-skip-reasons <number>             Limit skip reason rows in output (default: 10)',
  ].join('\n');
}

function expandHomePath(pathValue: string): string {
  if (!pathValue.startsWith('~')) {
    return pathValue;
  }
  const home = process.env.HOME;
  if (!home) {
    return pathValue;
  }
  return resolve(home, pathValue.slice(1));
}

function collapseHomePath(pathValue: string): string {
  const home = process.env.HOME;
  if (!home) {
    return pathValue;
  }
  return pathValue.startsWith(home) ? `~${pathValue.slice(home.length)}` : pathValue;
}

function parseTime(value: string | undefined, flag: string): number | undefined {
  if (!value) {
    return undefined;
  }
  if (/^\d+$/.test(value)) {
    const asNumber = Number(value);
    if (Number.isFinite(asNumber)) {
      return asNumber;
    }
  }
  const parsed = Date.parse(value);
  if (!Number.isFinite(parsed)) {
    throw new Error(`Invalid ${flag} value "${value}". Use ISO-8601 or epoch milliseconds.`);
  }
  return parsed;
}

function parseCsv(value: string | undefined): string[] | undefined {
  if (!value) {
    return undefined;
  }
  const values = value
    .split(',')
    .map((item) => item.trim())
    .filter((item) => item.length > 0);
  return values.length > 0 ? values : undefined;
}

function parseSources(raw: string | undefined): AuditSource[] | undefined {
  const values = parseCsv(raw);
  if (!values) {
    return undefined;
  }
  const parsed: AuditSource[] = [];
  for (const value of values) {
    if (value === 'gateway' || value === 'channel') {
      parsed.push(value);
      continue;
    }
    throw new Error(`Invalid source "${value}".`);
  }
  return parsed;
}

function parseOptionalNumber(raw: string | undefined, flag: string): number | undefined {
  if (!raw) {
    return undefined;
  }
  const parsed = Number(raw);
  if (!Number.isFinite(parsed)) {
    throw new Error(`Invalid ${flag} value "${raw}". Expected a number.`);
  }
  return parsed;
}

function isoDateTagNow(): string {
  return new Date().toISOString().slice(0, 10);
}

async function writeTextFile(pathValue: string, contents: string): Promise<void> {
  await mkdir(dirname(pathValue), { recursive: true });
  await writeFile(pathValue, contents, 'utf8');
}

async function main(): Promise<void> {
  const { values } = parseArgs({
    options: {
      audit: { type: 'string' },
      since: { type: 'string' },
      until: { type: 'string' },
      channel: { type: 'string' },
      source: { type: 'string' },
      'exclude-session-substring': { type: 'string' },
      'raw-identifiers': { type: 'boolean' },
      tag: { type: 'string' },
      'sample-out': { type: 'string' },
      'summary-json-out': { type: 'string' },
      'summary-md-out': { type: 'string' },
      'max-sessions': { type: 'string' },
      'max-channels': { type: 'string' },
      'max-skip-reasons': { type: 'string' },
      help: { type: 'boolean', short: 'h' },
    },
    strict: true,
    allowPositionals: false,
  });

  if (values.help) {
    process.stdout.write(`${usage()}\n`);
    return;
  }

  const auditPath = expandHomePath(values.audit ?? '~/.local/share/flynn/audit.log');
  const tag = values.tag ?? isoDateTagNow();
  const sampleOut = values['sample-out'] ?? `docs/plans/artifacts/phase0_baseline_live_${tag}.jsonl`;
  const summaryJsonOut = values['summary-json-out'] ?? `docs/plans/artifacts/phase0_baseline_live_${tag}.json`;
  const summaryMdOut = values['summary-md-out'] ?? `docs/plans/artifacts/phase0_baseline_live_${tag}.md`;
  const channels = parseCsv(values.channel);
  const sources = parseSources(values.source);
  const excludeSessionSubstrings = parseCsv(values['exclude-session-substring']) ?? ['probe'];
  const startTime = parseTime(values.since, '--since');
  const endTime = parseTime(values.until, '--until');

  const summaryOptions: Phase0BaselineSummaryOptions = {
    channels,
    sources,
    maxSessions: parseOptionalNumber(values['max-sessions'], '--max-sessions') ?? 20,
    maxChannels: parseOptionalNumber(values['max-channels'], '--max-channels') ?? 20,
    maxSkipReasons: parseOptionalNumber(values['max-skip-reasons'], '--max-skip-reasons') ?? 10,
  };

  const sourceEvents = await queryAuditLogs(auditPath, {
    start_time: startTime,
    end_time: endTime,
    event_types: [...DEFAULT_EVENT_TYPES],
  });

  const sampledEvents = capturePhase0LiveBaselineEvents(sourceEvents, {
    channels,
    sources,
    excludeSessionSubstrings,
    anonymizeIdentifiers: !values['raw-identifiers'],
  });

  const summary = summarizePhase0Baseline(sampledEvents, summaryOptions);
  const markdown = renderPhase0BaselineMarkdown(summary, summaryOptions);
  const sampleJsonl = sampledEvents.map((entry) => JSON.stringify(entry)).join('\n');
  const summaryJson = JSON.stringify({
    generated_at: new Date().toISOString(),
    source_audit_path: collapseHomePath(auditPath),
    source_event_count: sourceEvents.length,
    sampled_event_count: sampledEvents.length,
    filters: {
      since_ms: startTime,
      until_ms: endTime,
      channels,
      sources,
      exclude_session_substrings: excludeSessionSubstrings,
      anonymized_identifiers: !values['raw-identifiers'],
    },
    options: summaryOptions,
    summary,
  }, null, 2);

  await writeTextFile(sampleOut, sampleJsonl.length > 0 ? `${sampleJsonl}\n` : '');
  await writeTextFile(summaryJsonOut, `${summaryJson}\n`);
  await writeTextFile(summaryMdOut, `${markdown}\n`);

  process.stdout.write(`Captured ${sampledEvents.length} events from ${sourceEvents.length} source events.\n`);
  process.stdout.write(`- sample: ${sampleOut}\n`);
  process.stdout.write(`- summary json: ${summaryJsonOut}\n`);
  process.stdout.write(`- summary md: ${summaryMdOut}\n`);
}

main().catch((error) => {
  const message = error instanceof Error ? error.message : String(error);
  process.stderr.write(`${message}\n\n${usage()}\n`);
  process.exitCode = 1;
});