Files
flynn/scripts/capture-phase0-live-baseline.ts
T
William Valentin bfa857f074 fix(audit): dedupe repeated backend CLI selections
Normalize repeated --backend values in phase0 capture/drift scripts so backend lists are unique and deterministic. Architecture/protocol diagrams reviewed; no updates needed for this parsing-only change.
2026-02-27 13:27:35 -08:00

334 lines
12 KiB
JavaScript

#!/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 { normalizeArtifactTag } from '../src/audit/artifactTag.js';
import {
capturePhase0LiveBaselineEvents,
type Phase0BackendTarget,
} from '../src/audit/phase0LiveBaseline.js';
import { findLatestGatewayCancelWindow } from '../src/audit/phase0GatewayWindow.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;
const BACKEND_TARGETS: readonly Phase0BackendTarget[] = [
'native',
'claude_code',
'opencode',
'codex',
'gemini',
'pi_embedded',
];
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',
' --backend <native|pi_embedded|...[,..]> Restrict sample to selected backends (via backend.route timeline)',
' --exclude-session-substring <text[,..]> Exclude sessions containing any substring (default: probe)',
' --auto-gateway-cancel-window Auto-select latest gateway cancel/cancelled session window',
' --window-padding-ms <integer> Milliseconds added before/after auto-selected window (default: 250)',
' --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 <integer> Limit session rows in output (default: 20)',
' --max-channels <integer> Limit channel rows in output (default: 20)',
' --max-skip-reasons <integer> 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 parseOptionalInteger(raw: string | undefined, flag: string): number | undefined {
const parsed = parseOptionalNumber(raw, flag);
if (parsed === undefined) {
return undefined;
}
if (!Number.isInteger(parsed)) {
throw new Error(`Invalid ${flag} value "${raw}". Expected an integer.`);
}
if (parsed < 0) {
throw new Error(`${flag} must be greater than or equal to 0.`);
}
return parsed;
}
function parseBackendTargets(raw: string | undefined): Phase0BackendTarget[] | undefined {
const values = parseCsv(raw);
if (!values) {
return undefined;
}
const parsed: Phase0BackendTarget[] = [];
const seen = new Set<Phase0BackendTarget>();
for (const value of values) {
if (BACKEND_TARGETS.includes(value as Phase0BackendTarget)) {
const backend = value as Phase0BackendTarget;
if (!seen.has(backend)) {
parsed.push(backend);
seen.add(backend);
}
continue;
}
throw new Error(`Invalid backend "${value}".`);
}
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' },
backend: { type: 'string' },
'exclude-session-substring': { type: 'string' },
'auto-gateway-cancel-window': { type: 'boolean' },
'window-padding-ms': { 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(`${usage()}\n`);
return;
}
const auditPath = expandHomePath(values.audit ?? '~/.local/share/flynn/audit.log');
const tag = normalizeArtifactTag(values.tag ?? isoDateTagNow(), '--tag');
const channels = parseCsv(values.channel);
let sources = parseSources(values.source);
const backendTargets = parseBackendTargets(values.backend);
const excludeSessionSubstrings = parseCsv(values['exclude-session-substring']) ?? ['probe'];
const autoGatewayCancelWindow = Boolean(values['auto-gateway-cancel-window']);
const windowPaddingMs = parseOptionalInteger(values['window-padding-ms'], '--window-padding-ms');
if (autoGatewayCancelWindow && (values.since || values.until)) {
throw new Error('--auto-gateway-cancel-window cannot be combined with --since/--until.');
}
if (autoGatewayCancelWindow && sources && !sources.includes('gateway')) {
throw new Error('--auto-gateway-cancel-window requires --source to include "gateway" (or omit --source).');
}
let startTime = parseTime(values.since, '--since');
let endTime = parseTime(values.until, '--until');
let autoWindow: ReturnType<typeof findLatestGatewayCancelWindow> = null;
if (autoGatewayCancelWindow) {
sources = sources ?? ['gateway'];
const autoWindowSourceEvents = await queryAuditLogs(auditPath, {
event_types: ['run.state', 'run.cancel'],
});
autoWindow = findLatestGatewayCancelWindow(autoWindowSourceEvents, {
padding_ms: windowPaddingMs ?? 250,
});
if (!autoWindow) {
throw new Error('No gateway cancel/cancelled session window found in audit log.');
}
startTime = autoWindow.start_time_ms;
endTime = autoWindow.end_time_ms;
}
const isGatewayOnly = sources?.length === 1 && sources[0] === 'gateway';
const backendSuffix = backendTargets && backendTargets.length > 0
? backendTargets.length === 1
? `backend_${backendTargets[0]}`
: 'backend_scoped'
: undefined;
const defaultBaseName = isGatewayOnly
? `docs/plans/artifacts/phase0_baseline_live_gateway_${tag}`
: backendSuffix
? `docs/plans/artifacts/phase0_baseline_live_${backendSuffix}_${tag}`
: `docs/plans/artifacts/phase0_baseline_live_${tag}`;
const sampleOut = values['sample-out'] ?? `${defaultBaseName}.jsonl`;
const summaryJsonOut = values['summary-json-out'] ?? `${defaultBaseName}.json`;
const summaryMdOut = values['summary-md-out'] ?? `${defaultBaseName}.md`;
const summaryOptions: Phase0BaselineSummaryOptions = {
channels,
sources,
maxSessions: parseOptionalInteger(values['max-sessions'], '--max-sessions') ?? 20,
maxChannels: parseOptionalInteger(values['max-channels'], '--max-channels') ?? 20,
maxSkipReasons: parseOptionalInteger(values['max-skip-reasons'], '--max-skip-reasons') ?? 10,
};
const backendRouteEvents = backendTargets && backendTargets.length > 0
? await queryAuditLogs(auditPath, {
start_time: startTime,
end_time: endTime,
event_types: ['backend.route'],
})
: [];
const sourceEvents = await queryAuditLogs(auditPath, {
start_time: startTime,
end_time: endTime,
event_types: [...DEFAULT_EVENT_TYPES],
});
const sampledEvents = capturePhase0LiveBaselineEvents(sourceEvents, {
channels,
sources,
backendTargets,
backendRouteEvents,
excludeSessionSubstrings,
anonymizeIdentifiers: !values['raw-identifiers'],
});
const summary = summarizePhase0Baseline(sampledEvents, summaryOptions);
const markdown = renderPhase0BaselineMarkdown(summary, summaryOptions);
const sampleJsonl = sampledEvents.map((entry) => JSON.stringify(entry)).join('\n');
const summaryJson = JSON.stringify({
generated_at: new Date().toISOString(),
source_audit_path: collapseHomePath(auditPath),
source_event_count: sourceEvents.length,
sampled_event_count: sampledEvents.length,
filters: {
since_ms: startTime,
until_ms: endTime,
channels,
sources,
backend_targets: backendTargets,
exclude_session_substrings: excludeSessionSubstrings,
anonymized_identifiers: !values['raw-identifiers'],
auto_gateway_cancel_window: autoWindow
? {
...autoWindow,
padding_ms: windowPaddingMs ?? 250,
}
: undefined,
backend_route_event_count: backendRouteEvents.length > 0 ? backendRouteEvents.length : undefined,
},
options: summaryOptions,
summary,
}, null, 2);
await writeTextFile(sampleOut, sampleJsonl.length > 0 ? `${sampleJsonl}\n` : '');
await writeTextFile(summaryJsonOut, `${summaryJson}\n`);
await writeTextFile(summaryMdOut, `${markdown}\n`);
process.stdout.write(`Captured ${sampledEvents.length} events from ${sourceEvents.length} source events.\n`);
if (autoWindow) {
process.stdout.write(`- auto gateway window: session=${autoWindow.session_id} start=${autoWindow.start_time_ms} end=${autoWindow.end_time_ms}\n`);
}
if (backendTargets && backendTargets.length > 0) {
process.stdout.write(`- backend targets: ${backendTargets.join(', ')} (route events: ${backendRouteEvents.length})\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;
});