krow_backend/scripts/oracle.mjs

/**
 * The JS parser, as an oracle.
 *
 * Runs the REAL frontend module graph through Vite — `import.meta.glob`, the
 * `@/` alias and raw Markdown loading behave exactly as they do in the app, the
 * same technique the frontend's own `scripts/skill-check.mjs` uses (that file
 * lives in krow-demo, not here). A mock of the registry would reproduce none of
 * the behaviour this file exists to capture.
 *
 * Emits one JSON document: for every shipped definition and every adversarial
 * case, what the JS parser did with it. That document is the fixture the Go
 * conformance suite asserts against, so "the Go parser agrees with the JS
 * parser" is a comparison against the JS parser's actual output rather than
 * against anybody's description of it.
 */
import { readFileSync, readdirSync, writeFileSync, statSync } from 'node:fs';
import { join, relative } from 'node:path';

/* The frontend checkout. Overridable so this runs anywhere the two repos are
   checked out side by side, which is the layout it defaults to. */
const FRONTEND = process.env.KROW_FRONTEND
  || new URL('../../krow-demo', import.meta.url).pathname;
const { createServer } = await import(join(FRONTEND, 'node_modules/vite/dist/node/index.js'));
const { CASES } = await import(new URL('./cases.mjs', import.meta.url).href);

const server = await createServer({
  root: FRONTEND, server: { middlewareMode: true }, appType: 'custom', logLevel: 'error',
});

const skillReg = await server.ssrLoadModule('/src/lib/skills/registry.js');
const agentReg = await server.ssrLoadModule('/src/lib/agents/registry.js');

/** Every .md under a directory, recursively, repo-relative. */
function walk(dir) {
  const out = [];
  for (const entry of readdirSync(dir)) {
    const full = join(dir, entry);
    if (statSync(full).isDirectory()) out.push(...walk(full));
    else if (entry.endsWith('.md')) out.push(full);
  }
  return out.sort();
}

/** The fields the backend contract actually projects out of a definition. */
const projectSkill = (s) => ({
  id: s.id,
  name: s.name,
  description: s.description,
  status: s.status,
  pages: s.pages,
  kind: s.kind,
  category: s.category,
  actions: s.actions,
  triggers: s.triggers,
  declaredTriggers: s.declaredTriggers,
  prompt: s.prompt ?? null,
  facets: s.facets,
  skillId: s.skillId ?? null,
});

const projectAgent = (a) => ({
  id: a.id,
  name: a.name,
  description: a.description,
  status: a.status,
  version: a.version,
  pages: a.pages,
  icon: a.icon,
  reasoning: a.reasoning,
  trigger: a.trigger,
  webSearch: a.webSearch,
  skills: a.skills,
  subagents: a.subagents,
  starters: a.starters,
  permissions: a.permissions,
  errors: a.errors,
});

/** One definition, as the JS side sees it end to end. */
function observe(raw, kind) {
  const out = { kind };

  /* Layer 1 — the fence. */
  try {
    out.hasFrontmatter = skillReg.hasFrontmatter(raw);
    const { data, body } = skillReg.parseFrontmatter(raw);
    out.frontmatter = { ok: true, data, body };
  } catch (error) {
    out.hasFrontmatter = (() => { try { return skillReg.hasFrontmatter(raw); } catch { return null; } })();
    out.frontmatter = { ok: false, error: String(error?.message ?? error) };
  }

  /* Layer 2 — the definition. */
  try {
    const parsed = kind === 'agent'
      ? agentReg.parseAgent(raw, { custom: true })
      : skillReg.parseSkill(raw, { custom: true });
    out.parse = { ok: true };
    out.normalized = kind === 'agent' ? projectAgent(parsed) : projectSkill(parsed);
    out.markdownVerbatim = parsed.markdown === raw;
  } catch (error) {
    out.parse = { ok: false, error: String(error?.message ?? error) };
    out.normalized = null;
    out.markdownVerbatim = null;
  }

  /* Layer 3 — the save gate. This is the accept/reject contract. */
  const problem = kind === 'agent'
    ? agentReg.validateAgentSource(raw)
    : skillReg.validateSkillSource(raw);
  out.accepted = problem === null;
  out.rejection = problem;

  return out;
}

/* ── The shipped corpus ───────────────────────────────────────────────────── */
const corpus = [];
const groups = [
  { dir: join(FRONTEND, 'src/agents'), type: 'agent', kind: 'agent' },
  { dir: join(FRONTEND, 'src/skills'), type: 'skill', kind: 'skill' },
  { dir: join(FRONTEND, 'skill-examples'), type: 'example', kind: 'skill' },
];

for (const { dir, type, kind } of groups) {
  for (const file of walk(dir)) {
    const raw = readFileSync(file, 'utf8');
    corpus.push({
      path: relative(FRONTEND, file),
      type,
      rawBase64: Buffer.from(raw, 'utf8').toString('base64'),
      bytes: Buffer.byteLength(raw, 'utf8'),
      ...observe(raw, kind),
    });
  }
}

/* ── The adversarial cases ────────────────────────────────────────────────── */
const cases = CASES.map((c) => ({
  name: c.name,
  rawBase64: Buffer.from(c.raw, 'utf8').toString('base64'),
  bytes: Buffer.byteLength(c.raw, 'utf8'),
  ...observe(c.raw, c.kind),
}));

/* ── The closed vocabularies, read out of the JS tables themselves ────────── */
const surfaces = await server.ssrLoadModule('/src/lib/skills/surfaces.js');
const agentVocab = await server.ssrLoadModule('/src/lib/agents/vocabulary.js');

const vocabulary = {
  pages: surfaces.SKILL_SURFACES.map((s) => ({ id: s.id, aliases: s.aliases || [] })),
  agentStatuses: agentVocab.AGENT_STATUSES,
  defaultAgentStatus: agentVocab.DEFAULT_AGENT_STATUS,
  reasoning: agentVocab.SUPPORTED_REASONING,
  defaultReasoning: agentVocab.DEFAULT_REASONING,
  icons: agentVocab.AGENT_ICONS,
  defaultIcon: agentVocab.DEFAULT_AGENT_ICON,
  knowledgeKinds: agentVocab.KNOWLEDGE_KINDS,
  defaultKnowledgeKind: agentVocab.DEFAULT_KNOWLEDGE_KIND,
  access: agentVocab.AGENT_ACCESS,
  defaultAccess: agentVocab.DEFAULT_AGENT_ACCESS,
  roles: agentVocab.PERMISSION_ROLES,
  defaultRole: agentVocab.DEFAULT_PERMISSION_ROLE,
};

await server.close();

const doc = {
  generatedBy: 'scripts/oracle.mjs against the frontend module graph',
  frontendParser: {
    yaml: 'src/lib/skills/yaml.js (hand-written YAML subset, no dependency)',
    frontmatter: 'src/lib/skills/registry.js — normalizeDefinition / hasFrontmatter / parseFrontmatter',
    skill: 'src/lib/skills/registry.js — parseSkill / validateSkillSource',
    agent: 'src/lib/agents/registry.js — parseAgent / validateAgentSource',
  },
  vocabulary,
  corpus,
  cases,
};

const target = process.argv[2];
writeFileSync(target, JSON.stringify(doc, null, 2) + '\n');

const acc = (xs) => xs.filter((x) => x.accepted).length;
console.log(`corpus ${corpus.length} files — ${acc(corpus)} accepted, ${corpus.length - acc(corpus)} rejected`);
console.log(`cases  ${cases.length} — ${acc(cases)} accepted, ${cases.length - acc(cases)} rejected`);
console.log(`markdown verbatim: ${[...corpus, ...cases].every((x) => x.markdownVerbatim !== false)}`);
console.log(`written: ${target}`);