189 lines
6.9 KiB
JavaScript
189 lines
6.9 KiB
JavaScript
#!/usr/bin/env node
|
|
// audit-corpus-headers.mjs — Spor 1 Step 1: read-only audit of the corpus header state.
|
|
//
|
|
// Measures, per non-advisor reference file and in aggregate, the empirical ground
|
|
// truth the migration plan rests on: header-region **Verified:** values (clean date
|
|
// vs stale "MCP <date>"), body-duplicates and pipe-delimited meta rows inside the
|
|
// 500-byte header window, base-field presence (English **Last updated:**, **Status:**,
|
|
// title) and the header dialect (**Category:** / **Kategori:** / pipe). Mutates nothing.
|
|
//
|
|
// Usage:
|
|
// node scripts/kb-update/audit-corpus-headers.mjs # human summary
|
|
// node scripts/kb-update/audit-corpus-headers.mjs --json # full report to stdout
|
|
// node scripts/kb-update/audit-corpus-headers.mjs --write # persist data/corpus-header-audit.json
|
|
|
|
import { readdirSync, readFileSync, existsSync, realpathSync } from 'node:fs';
|
|
import { join, relative, dirname } from 'node:path';
|
|
import { fileURLToPath } from 'node:url';
|
|
import { atomicWriteJson } from './lib/atomic-write.mjs';
|
|
|
|
const __dirname = dirname(fileURLToPath(import.meta.url));
|
|
const PLUGIN_ROOT = join(__dirname, '..', '..');
|
|
const SKILLS_DIR = join(PLUGIN_ROOT, 'skills');
|
|
const OUT_PATH = join(__dirname, 'data', 'corpus-header-audit.json');
|
|
|
|
// The four non-advisor skills — advisor is out of audit scope (S-Cosmo: never mutated,
|
|
// and its header state is irrelevant to the Step-5/9 blast radius this audit pins).
|
|
const NON_ADVISOR_SKILLS = [
|
|
'ms-ai-engineering',
|
|
'ms-ai-governance',
|
|
'ms-ai-infrastructure',
|
|
'ms-ai-security',
|
|
];
|
|
|
|
// Mirrors kb-headers.mjs — the Port-1 contract reads only the top 500 bytes.
|
|
const HEADER_REGION_BYTES = 500;
|
|
|
|
// Base-field regexes mirror transform.mjs validateKbFile (RE_LAST_UPDATED:261 is not
|
|
// exported; replicated verbatim so the audit measures exactly what the guard reads).
|
|
const RE_TITLE = /^#\s+\S/m;
|
|
const RE_LAST_UPDATED = /\*\*Last updated:\*\*\s*[\d-]+/i;
|
|
const RE_STATUS = /\*\*Status:\*\*\s*\S/i;
|
|
|
|
// Verified value = everything after the label up to a pipe separator or line end,
|
|
// so "MCP 2026-06" survives intact while a pipe row stops at its sibling token.
|
|
const RE_VERIFIED = /\*\*Verified:\*\*\s*([^|\n]*)/;
|
|
const RE_CLEAN_DATE = /^\d{4}-\d{2}(-\d{2})?$/;
|
|
|
|
/** Audit a single file's content. Pure. */
|
|
function auditContent(content) {
|
|
const head = content.slice(0, HEADER_REGION_BYTES);
|
|
|
|
const verifiedMatches = [...head.matchAll(new RegExp(RE_VERIFIED.source, 'g'))];
|
|
let verified = null;
|
|
if (verifiedMatches.length > 0) {
|
|
const value = verifiedMatches[0][1].trim();
|
|
const line = head.slice(0, verifiedMatches[0].index).split('\n').length - 1;
|
|
const lineText = head.split('\n')[line] ?? '';
|
|
verified = {
|
|
value,
|
|
isDate: RE_CLEAN_DATE.test(value),
|
|
pipe: lineText.includes(' | '),
|
|
duplicateWithinHeaderRegion: verifiedMatches.length > 1,
|
|
};
|
|
}
|
|
|
|
let dialect = 'none';
|
|
if (/\*\*[^*\n]+:\*\*[^\n]* \| \*\*/.test(head)) dialect = 'pipe';
|
|
else if (/\*\*Category:\*\*/i.test(head)) dialect = 'category';
|
|
else if (/\*\*Kategori:\*\*/i.test(head)) dialect = 'kategori';
|
|
|
|
return {
|
|
title: RE_TITLE.test(content),
|
|
status: RE_STATUS.test(content),
|
|
lastUpdatedEnglish: RE_LAST_UPDATED.test(content),
|
|
dialect,
|
|
verified,
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Audit a batch of files. Pure + dependency-injected reader (validate-kb-file.mjs
|
|
* pattern) so tests never touch disk. An unreadable file is reported, not thrown.
|
|
*
|
|
* @param {string[]} paths
|
|
* @param {(p: string) => string} [readFile]
|
|
* @returns {{files: Record<string, object>, aggregate: object}}
|
|
*/
|
|
export function auditHeaders(paths, readFile = (p) => readFileSync(p, 'utf8')) {
|
|
const files = {};
|
|
const aggregate = {
|
|
files: 0,
|
|
readErrors: 0,
|
|
verifiedPresent: 0,
|
|
verifiedDate: 0,
|
|
verifiedNonDate: 0,
|
|
verifiedPipe: 0,
|
|
verifiedDuplicate: 0,
|
|
missingTitle: 0,
|
|
missingStatus: 0,
|
|
missingEnglishLastUpdated: 0,
|
|
dialects: { category: 0, kategori: 0, pipe: 0, none: 0 },
|
|
};
|
|
|
|
for (const path of paths ?? []) {
|
|
aggregate.files++;
|
|
let content;
|
|
try {
|
|
content = readFile(path);
|
|
} catch (err) {
|
|
files[path] = { error: err.message };
|
|
aggregate.readErrors++;
|
|
continue;
|
|
}
|
|
const r = auditContent(content);
|
|
files[path] = r;
|
|
if (r.verified) {
|
|
aggregate.verifiedPresent++;
|
|
if (r.verified.isDate) aggregate.verifiedDate++;
|
|
else aggregate.verifiedNonDate++;
|
|
if (r.verified.pipe) aggregate.verifiedPipe++;
|
|
if (r.verified.duplicateWithinHeaderRegion) aggregate.verifiedDuplicate++;
|
|
}
|
|
if (!r.title) aggregate.missingTitle++;
|
|
if (!r.status) aggregate.missingStatus++;
|
|
if (!r.lastUpdatedEnglish) aggregate.missingEnglishLastUpdated++;
|
|
aggregate.dialects[r.dialect]++;
|
|
}
|
|
|
|
return { files, aggregate };
|
|
}
|
|
|
|
// Walk directory recursively for .md files (build-registry.mjs shape).
|
|
function walkMd(dir) {
|
|
const results = [];
|
|
if (!existsSync(dir)) return results;
|
|
for (const entry of readdirSync(dir, { withFileTypes: true })) {
|
|
const full = join(dir, entry.name);
|
|
if (entry.isDirectory()) {
|
|
results.push(...walkMd(full));
|
|
} else if (entry.name.endsWith('.md') && entry.name !== 'SKILL.md') {
|
|
results.push(full);
|
|
}
|
|
}
|
|
return results;
|
|
}
|
|
|
|
function main(argv) {
|
|
const json = argv.includes('--json');
|
|
const write = argv.includes('--write');
|
|
|
|
const paths = NON_ADVISOR_SKILLS
|
|
.flatMap((skill) => walkMd(join(SKILLS_DIR, skill, 'references')))
|
|
.sort();
|
|
const report = auditHeaders(paths);
|
|
|
|
// Re-key on plugin-relative paths for a stable, machine-diffable artifact.
|
|
const relFiles = {};
|
|
for (const [abs, r] of Object.entries(report.files)) {
|
|
relFiles[relative(PLUGIN_ROOT, abs)] = r;
|
|
}
|
|
const out = { generated_by: 'audit-corpus-headers.mjs', files: relFiles, aggregate: report.aggregate };
|
|
|
|
if (write) {
|
|
atomicWriteJson(OUT_PATH, out);
|
|
console.log(`Wrote ${relative(PLUGIN_ROOT, OUT_PATH)}`);
|
|
}
|
|
if (json) {
|
|
process.stdout.write(JSON.stringify(out, null, 2) + '\n');
|
|
}
|
|
if (!json) {
|
|
const a = report.aggregate;
|
|
console.log(`Corpus header audit — ${a.files} non-advisor reference files`);
|
|
console.log(` Verified header present: ${a.verifiedPresent} (date: ${a.verifiedDate}, stale non-date: ${a.verifiedNonDate})`);
|
|
console.log(` Verified pipe rows: ${a.verifiedPipe}, duplicates within 500B: ${a.verifiedDuplicate}`);
|
|
console.log(` Missing English Last updated: ${a.missingEnglishLastUpdated}`);
|
|
console.log(` Missing Status: ${a.missingStatus}`);
|
|
console.log(` Missing title: ${a.missingTitle}`);
|
|
console.log(` Dialects: category=${a.dialects.category} kategori=${a.dialects.kategori} pipe=${a.dialects.pipe} none=${a.dialects.none}`);
|
|
}
|
|
}
|
|
|
|
const isMain = (() => {
|
|
try {
|
|
return realpathSync(process.argv[1]) === realpathSync(fileURLToPath(import.meta.url));
|
|
} catch {
|
|
return false;
|
|
}
|
|
})();
|
|
if (isMain) main(process.argv.slice(2));
|