msd-core/src/api-coverage.cts

/**
 * API-Coverage detector + matrix validator (#1562).
 *
 * The enforcement half of "Full API Coverage by Default — Opt Out, Never Opt In."
 * When a phase integrates an external API/service/SDK, the planner must produce a
 * coverage matrix (COVERAGE.md) enumerating the API's capability surface; every
 * non-integrated capability is an explicit, reasoned opt-out. The seal-time gate
 * (capabilities/ai-integration, verify:pre) consumes this module to (a) detect
 * whether a phase integrates an external API and (b) validate the produced matrix.
 *
 * Design notes (rubber-duck'd):
 *  - DETERMINISTIC + TYPED IR. Both the "does this phase integrate an external
 *    API?" decision and the "is this matrix complete?" decision are pure
 *    functions returning typed IR, not LLM judgments — so the low-false-positive
 *    guarantee (acceptance criterion #4) and the completeness guarantee
 *    (acceptance #2) are testable. Mirrors assumption-delta.cts (#1561).
 *  - COMPOUND SIGNAL for low false positives. A bare word like "api" appears in
 *    countless non-integration phases ("the public API of UserController"). The
 *    detector requires an INTEGRATION VERB and an EXTERNAL-API NOUN in the SAME
 *    CLAUSE (#2365 — same-line co-occurrence across unrelated clauses over-fired;
 *    the clause boundary, not a word-gap cap, is the relationship test), or an
 *    explicit "<Service> API/SDK" phrase naming a real service. Single weak
 *    tokens do not fire. This is the issue's "low false-positive trigger" made
 *    mechanical.
 *  - CODE AND PATHS ARE NOT PROSE. Fenced code blocks and inline code spans are
 *    stripped first (markdown-sectionizer seam), and path-shaped tokens
 *    (`src/app/api/...`, URLs) are masked, so a trigger term inside code or a
 *    first-party route path does not fire (#2365).
 *  - NO-INTEGRATION DECLARATION (#2365 acceptance #5). A COVERAGE.md consisting
 *    of `No external API integration: <reason>` is a valid, reasoned way for a
 *    phase to state that no external surface exists — the alternative to
 *    fabricating a matrix row when the detector is overruled by a human.
 *  - THE DETECTOR IS A FALLBACK. The primary path is the plan:pre contribution
 *    prompting COVERAGE.md creation. The detector runs only when COVERAGE.md is
 *    ABSENT, to catch the "nobody decided" case (acceptance #1). Its precision
 *    therefore matters but is not the only line of defense.
 *  - MATRIX FORMAT. The matrix is a markdown table (human-editable, diff-friendly)
 *    with a header row `| capability | decision | reason |` and one row per
 *    capability. decision ∈ {INTEGRATE, OPT-OUT}. An OPT-OUT row MUST carry a
 *    non-empty reason. A fenced ```coverage JSON block is also accepted for
 *    machine-generated matrices. This dual shape is bijective (parse/render
 *    round-trip) and covered by a fast-check property test.
 *  - ADDITIVE-ONLY VOCABULARY (Hyrum's Law). Once shipped, the verb/noun sets
 *    are depended-upon interfaces; they only grow. Tunable via the `terms`
 *    parameter so teams can widen them without forking.
 *
 * Public API:
 *   detectApiIntegration(text, terms?) -> { detected, signals, terms }
 *   parseCoverageMatrix(text) -> { rows, errors, format }
 *   validateCoverageMatrix(text) -> { valid, errors, counts }
 *   renderCoverageMatrix(rows) -> string
 *   DEFAULT_API_COVERAGE_TERMS
 *
 * CLI:
 *   echo "$SCOPE" | node gsd-core/bin/lib/api-coverage.cjs [--json]
 *     exit 0 = integration detected, 1 = none, 2 = startup error
 */

import { stripFencedCode, scanInlineCodeSpans, extractFencedBlock } from './markdown-sectionizer.cjs';
import { escapeRegex } from './pattern.cjs';

// ─── Integration-signal vocabulary ────────────────────────────────────────────

export interface ApiCoverageTermSet {
  verbs: string[];
  nouns: string[];
}

export interface ApiCoverageSignal {
  verb: string;
  noun: string;
  snippet: string;
}

export interface ApiCoverageDetectionResult {
  detected: boolean;
  signals: ApiCoverageSignal[];
  terms: ApiCoverageTermSet;
}

/**
 * Curated default trigger vocabulary. ADDITIVE-ONLY (Hyrum's Law). Tunable via
 * the `terms` parameter.
 *
 * VERBS are deliberately conservative: common verbs like "add", "use", "call",
 * "implement" are EXCLUDED because they appear in nearly every phase and would
 * make the gate fire on prose that has nothing to do with an external API. The
 * verbs kept all connote BRINGING IN an external surface.
 *
 * NOUNS name an external-API surface. Bare "client" is excluded — too ambiguous
 * (client-side UI vs API client). "service" alone is excluded (internal
 * services); a phase integrating an external service virtually always pairs it
 * with "API"/"SDK"/"REST"/etc., which the compound verb+noun rule captures.
 */
export const DEFAULT_API_COVERAGE_TERMS: Readonly<ApiCoverageTermSet> = {
  verbs: [
    'integrate',
    'integrates',
    'integrating',
    'integration',
    'wrap',
    'wraps',
    'wrapping',
    'connect',
    'connects',
    'connecting',
    'consume',
    'consumes',
    'consuming',
    'wire',
    'wires',
    'wiring',
    'onboard',
    'onboarding',
    'adopt',
    'adopts',
    'adopting',
  ],
  nouns: [
    'api',
    'apis',
    'sdk',
    'sdks',
    'rest',
    'graphql',
    'grpc',
    'endpoint',
    'endpoints',
    'oauth',
    'oauth2',
    'webhook',
    'webhooks',
    'mcp',
  ],
};

/** Hardening caps for the tunable vocabulary (hostile `--terms` defense). */
const MAX_TERMS_PER_KIND = 200;
const MAX_TERM_LEN = 32;

/**
 * Field-length caps for matrix cell values. Cell content flows from a
 * semi-trusted COVERAGE.md into the gate `message` that the orchestrator LLM
 * reads, so it is bounded to keep the prompt-injection surface small and to
 * document the format contract (short, single-line prose — not paragraphs).
 */
const CAPABILITY_MAX_LEN = 80;
const REASON_MAX_LEN = 200;

function normalizeTerms(list: unknown): string[] {
  if (!Array.isArray(list)) return [];
  const seen = new Set<string>();
  const out: string[] = [];
  for (const raw of list) {
    if (typeof raw !== 'string') continue;
    const t = raw.trim().toLowerCase().slice(0, MAX_TERM_LEN);
    if (!t || !/[a-z0-9]/.test(t)) continue;
    if (seen.has(t)) continue;
    seen.add(t);
    out.push(t);
    if (out.length >= MAX_TERMS_PER_KIND) break;
  }
  return out;
}

function resolveTerms(terms?: Partial<ApiCoverageTermSet>): ApiCoverageTermSet {
  const merge = (key: 'verbs' | 'nouns'): string[] => {
    const t = terms && terms[key];
    return Array.isArray(t) ? normalizeTerms(t) : [...DEFAULT_API_COVERAGE_TERMS[key]];
  };
  return { verbs: merge('verbs'), nouns: merge('nouns') };
}

function makeSnippet(line: string, anchor: string): string {
  const cleaned = line.replace(/\s+/g, ' ').trim();
  if (cleaned.length <= 120) return cleaned;
  const idx = cleaned.toLowerCase().indexOf(anchor);
  if (idx < 0) return cleaned.slice(0, 120);
  const start = Math.max(0, idx - 50);
  const end = Math.min(cleaned.length, idx + anchor.length + 50);
  const prefix = start > 0 ? '…' : '';
  const suffix = end < cleaned.length ? '…' : '';
  return `${prefix}${cleaned.slice(start, end)}${suffix}`;
}

/** `<Service> API` / `<Service> SDK` — a capitalized proper noun immediately
 *  followed by API/SDK. Strong signal on its own (no verb required).
 *
 *  STOPWORDS guard against the false positive where an ordinary capitalized
 *  sentence starter ("The API …", "An SDK …", "Our REST …") matches the
 *  `[A-Z]\w+ API` shape. Those are common English, not a service name, so they
 *  are rejected before counting as a surface signal (acceptance #4 — low false
 *  positives). */
// Service-name length is bounded ({1,40}) so a hostile "A-A-A-…-A-x" run cannot
// drive the greedy group into O(n^2) backtracking (#2365 review). Nearly all
// vendor names fit; a >41-char service token before API/SDK would be missed by
// this surface path (it would still fire via the compound verb+noun rule) —
// an accepted bound.
const SERVICE_SURFACE_API_RE = /\b([A-Z][A-Za-z0-9_-]{1,40})\s+(API|SDK|REST|GraphQL)\b/;
const SERVICE_STOPWORDS = new Set([
  'the', 'an', 'a', 'our', 'this', 'these', 'that', 'those', 'new', 'add',
  'use', 'your', 'my', 'no', 'some', 'any', 'all', 'each', 'every', 'both',
  'if', 'when', 'while', 'with', 'via', 'using', 'into', 'its', 'their',
  'we', 'you', 'they', 'it',
]);

/** #2365 — the detector is FAIL-CLOSED: it leans toward detecting, because a
 *  false positive is cheaply dismissed by a one-line COVERAGE.md "no external
 *  API integration" declaration, whereas a false NEGATIVE silently lets a real
 *  external-API phase past a BLOCKING gate. So the only prose the detector
 *  actively suppresses is the classes that are unambiguously NOT external
 *  integration: first-party route paths, verb/noun in unrelated clauses, and
 *  descriptive/protocol "<Word> API" prose with no named service.
 *
 *  CLAUSE_BOUNDARY_RE: a verb and a noun form ONE compound action only inside
 *  one grammatical clause — sentence punctuation and table-cell walls (`|`)
 *  end a clause. `-` is deliberately absent (it would split hyphenated words).
 *  There is deliberately NO word-gap cap inside a clause: a cap cannot separate
 *  a genuine long integration clause (F4, 21 words) from a long internal-UI
 *  clause (18 words) — the clause boundary is the only sound signal, and the
 *  declaration handles the residual false positives. */
const CLAUSE_BOUNDARY_RE = /[,;:.!?|()—–]/;
/** Same character class as CLAUSE_BOUNDARY_RE, as a set — for scanning a token's
 *  trailing punctuation without an unanchored `[…]+$` regex, whose backtracking
 *  is O(n^2) on a long punctuation run (#2365 review). */
const CLAUSE_BOUNDARY_CHARS = new Set([',', ';', ':', '.', '!', '?', '|', '(', ')', '—', '–']);

/*  DELIBERATELY NO cross-clause binding. Detection is same-clause only. Binding
 *  a verb in one clause to a noun in another ("Integrate Stripe, exposing its
 *  endpoints"; "Integrate Stripe; use its endpoints") requires knowing "Stripe"
 *  is a vendor and "its" refers to it — a vendor dictionary + coreference, which
 *  trek-e's brief rules out in principle. Every lexical cross-clause rule tried
 *  (word-gap cap, participle continuation) traded a false negative for a false
 *  positive across four review rounds. So a service named ONLY in a clause
 *  separate from its API noun, with no explicit `<Service> API` surface, is a
 *  DOCUMENTED fail-open limitation — cheaply covered by the COVERAGE.md
 *  declaration and rare in real phase prose, which says "integrate the X API". */

/** In the `<Service> API|SDK` surface position, these capture words are NOT a
 *  named third-party service: locality/scope descriptors ("Internal API",
 *  "Public API") and bare protocol names ("REST API", "GraphQL API"). A real
 *  vendor name (Stripe, Shopify) is none of these, so rejecting them costs no
 *  true positives while killing the descriptive-prose false positives (#2365
 *  acceptance #3, review F8). */
const SURFACE_DESCRIPTOR_WORDS = new Set([
  'internal', 'external', 'public', 'private', 'local', 'in-house', 'first-party',
  'generic', 'shared', 'common', 'legacy', 'rest', 'restful', 'graphql', 'grpc',
  'soap', 'rpc', 'http', 'https', 'json', 'xml',
]);

/** Locality qualifiers that, when they immediately precede a `<Service> API`,
 *  mark it as first-party ("internal Payments API") — negative evidence for an
 *  EXTERNAL-API surface signal. Only unambiguously-internal words: "external"
 *  is deliberately absent (an external API IS external). */
const INTERNAL_DESCRIPTORS = new Set(['internal', 'in-house', 'local', 'first-party', 'private']);

/** #2784: negation suppression. A clause that pairs an integration verb with an
 *  API noun but the verb itself is directly negated (e.g. "does not integrate",
 *  "integrates no external API") is suppressed. Two windows are checked: a
 *  negation qualifier within 2 words directly before the verb, or "no"/"zero"/
 *  "none" between the verb and a following noun.
 *  KNOWN LIMIT (deliberate, not a bug to fix later): a negation further than 2
 *  words before the verb, or a clause where the noun precedes the verb, is NOT
 *  suppressed — e.g. "Ships without any API integration." still reports
 *  detected:true, because "without" sits outside the verb's 2-word lookback
 *  and the noun precedes the verb. This is intentional: detectApiIntegration
 *  is fail-closed — an unsuppressed false positive costs a one-line
 *  COVERAGE.md declaration, while widening the suppression window risks a
 *  false negative that silently lets a real external-API phase past a
 *  blocking gate.
 *  Hoisted to module scope (#3127 regression fix): this was previously
 *  allocated fresh on every source line, which is wasted work on documents
 *  with many lines. */
const NEGATION_QUALIFIERS = new Set([
  'no', 'not', 'without', 'zero', 'neither', 'nor', 'none', "don't", "doesn't", "didn't", "won't", "can't", "cannot",
]);
/** Negation tokens checked BETWEEN a verb and a later noun (narrower than
 *  NEGATION_QUALIFIERS — "not" and "without" are checked only immediately
 *  before the verb, via NEGATION_QUALIFIERS above). */
const NEGATION_NOUN_TOKENS = new Set(['no', 'zero', 'none']);

/** A capitalized compound modifier ("Resolver-only", "Read-only", "E-commerce"
 *  — lowercase letter right after the hyphen) is an adjective phrase, not a
 *  service name. Real hyphenated services capitalize the second segment
 *  ("T-Mobile"). */
const COMPOUND_MODIFIER_RE = /^[A-Z][A-Za-z0-9]*-[a-z]/;

interface TermMatch {
  term: string;
  start: number;
  end: number;
}

interface LineScan {
  /** The line with path-shaped tokens replaced by same-length space padding
   *  (offsets preserved for the clause logic). */
  masked: string;
  /** Noun-vocabulary terms found inside NON-LOCAL URLs (`https://api.stripe.com`)
   *  — a URL that itself names an API surface is external-dependency evidence,
   *  so it still feeds the compound rule even though the URL is masked from
   *  plain prose matching. */
  urlNouns: TermMatch[];
}

const URL_TOKEN_RE = /^[([<"'`]*[a-z][a-z0-9+.-]*:\/\//i;
const LOCAL_URL_RE = /^[([<"'`]*[a-z][a-z0-9+.-]*:\/\/(?:localhost|127(?:\.\d{1,3}){1,3}|0\.0\.0\.0|\[::1\])(?=[:/?#]|$)/i;
/** A scheme-less token that STARTS with a dotted hostname whose final label is
 *  alphabetic ("api.stripe.com/v1") — a bare external API host. A first-party
 *  route path ("src/app/api/…") has no dotted head, and an IP host ("127.1/…")
 *  has a numeric final label, so neither matches (#2365 review F2). */
const DOMAIN_HEAD_RE = /^[([<"'`]*(?:[a-z0-9](?:[a-z0-9-]*[a-z0-9])?\.)+[a-z]{2,}(?=[:/?#]|$)/i;

/** Mask whitespace-delimited tokens with an interior `/` — file paths, framework
 *  routes (`src/app/api/...`), URLs. They are references, not integration prose
 *  (#2365 root cause 2: `/` counted as a word boundary, so first-party route
 *  paths matched the noun vocabulary). Two carve-outs keep genuine signals:
 *   - a slashed token whose segments are ALL noun-vocabulary words ("API/SDK",
 *     "REST/GraphQL") is prose shorthand, not a path — left unmasked;
 *   - a non-local URL is masked, but noun terms inside it are collected as
 *     compound-rule evidence (the old detector caught "connect to
 *     https://api.stripe.com" via the `api` segment; losing that would
 *     fail-open). */
function scanLineTokens(line: string, nounRe: RegExp | null, nounSet: Set<string>): LineScan {
  const urlNouns: TermMatch[] = [];
  let masked = '';
  const tokenRe = /\S+/g;
  let last = 0;
  let m: RegExpExecArray | null;
  while ((m = tokenRe.exec(line)) !== null) {
    const rawTok = m[0];
    masked += line.slice(last, m.index);
    last = m.index + rawTok.length;
    // Peel trailing clause-boundary punctuation off the token and keep it
    // LITERAL in `masked` — masking it away would erase a clause split and pair
    // unrelated verb/noun across it (#2365 review F6: "…example.com, document…").
    // A backward char scan (not a `[…]+$` regex) keeps this linear.
    let trailLen = 0;
    while (trailLen < rawTok.length && CLAUSE_BOUNDARY_CHARS.has(rawTok[rawTok.length - 1 - trailLen])) {
      trailLen++;
    }
    const trail = trailLen ? rawTok.slice(rawTok.length - trailLen) : '';
    const tok = trailLen ? rawTok.slice(0, rawTok.length - trailLen) : rawTok;
    if (!/\S[\\/]\S/.test(tok)) {
      masked += rawTok;
      continue;
    }
    const segments = tok.split(/[\\/]/).map((s) => s.replace(/[^A-Za-z0-9]/g, ''));
    if (
      segments.every((s) => s.length > 0 && (nounSet.has(s.toLowerCase()) || /^v\d+$/i.test(s))) &&
      segments.some((s) => nounSet.has(s.toLowerCase()))
    ) {
      masked += rawTok; // "API/SDK", "API/v2" — noun shorthand, not a path
      continue;
    }
    // A scheme URL or a bare external hostname is an external dependency
    // reference: mask it from prose but keep it as compound-rule evidence. A
    // first-party route path has neither a scheme nor a dotted host, so it is
    // masked WITHOUT contributing nouns (#2365 root cause 2).
    // A non-local URL that NAMES an API vocabulary word ("api.stripe.com/v1")
    // is external-dependency evidence, so its vocab nouns feed the compound
    // rule. We deliberately do NOT treat every path-bearing URL as an endpoint:
    // that fired on ordinary asset/link URLs ("…/theme.css", "…?next=/x") and
    // recreated routine UI-phase false positives (#2365 review). A bare external
    // host that names no vocabulary word ("graph.microsoft.com") and is not
    // written as "<Service> API" is therefore a DOCUMENTED fail-open limitation.
    const isSchemeUrl = URL_TOKEN_RE.test(tok) && !LOCAL_URL_RE.test(tok);
    const isDomainUrl = !URL_TOKEN_RE.test(tok) && DOMAIN_HEAD_RE.test(tok);
    if (nounRe && (isSchemeUrl || isDomainUrl)) {
      for (const f of collectTermMatches(nounRe, tok)) {
        urlNouns.push({ term: f.term, start: m.index, end: m.index + tok.length });
      }
    }
    masked += ' '.repeat(tok.length) + trail;
  }
  masked += line.slice(last);
  return { masked, urlNouns };
}

/** All term matches in a clause, with offsets. `re` must be global with the
 *  term in group 2 and a consumed leading boundary in group 1. */
function collectTermMatches(re: RegExp, clause: string): TermMatch[] {
  const out: TermMatch[] = [];
  re.lastIndex = 0;
  let m: RegExpExecArray | null;
  while ((m = re.exec(clause)) !== null) {
    const start = m.index + (m[1] || '').length;
    out.push({ term: (m[2] || '').toLowerCase(), start, end: start + (m[2] || '').length });
    if (m[0].length === 0) re.lastIndex++;
  }
  return out;
}

interface ClauseSpan {
  text: string;
  start: number;
}

/** Split a line into clause segments, keeping each segment's start offset so
 *  line-level spans (masked URL tokens) can be mapped into their clause. */
function splitClauses(masked: string): ClauseSpan[] {
  const out: ClauseSpan[] = [];
  let start = 0;
  for (let i = 0; i <= masked.length; i++) {
    if (i === masked.length || CLAUSE_BOUNDARY_RE.test(masked[i])) {
      out.push({ text: masked.slice(start, i), start });
      start = i + 1;
    }
  }
  return out;
}

/**
 * Detect whether phase-scope prose describes integrating an external API/SDK.
 *
 * FAIL-CLOSED: it leans toward detecting, because a false positive is dismissed
 * by a one-line COVERAGE.md declaration while a false negative silently slips a
 * real external-API phase past a blocking gate. It fires when EITHER:
 *   (a) an integration VERB and an API NOUN share one CLAUSE ("integrate the
 *       Stripe API", "Connect … to api.stripe.com") — the clause boundary is the
 *       whole relationship test, so verb/noun in DIFFERENT clauses do not pair
 *       (#2365 acceptance #2). There is NO cross-clause binding: a service named
 *       only in a clause separate from its API noun is a documented limitation.
 *   (b) an explicit `<Service> API|SDK|REST|GraphQL` surface names a service
 *       that is not a stopword, a locality/protocol descriptor, a compound
 *       modifier, or first-party-qualified ("Stripe API", "Spotify SDK").
 *
 * Fenced code, inline code spans, and path-shaped tokens are excluded before
 * matching. A package-shaped inline span (`@stripe/stripe-js`, `stripe-sdk`)
 * and a URL that NAMES an API vocab word ("api.stripe.com/v1") still count as
 * noun/dependency evidence; a bare host that names none does not.
 *
 * Non-string inputs degrade to `{ detected: false }` without throwing.
 */
export function detectApiIntegration(
  text: unknown,
  terms?: Partial<ApiCoverageTermSet>,
): ApiCoverageDetectionResult {
  const effective = resolveTerms(terms);
  if (typeof text !== 'string') {
    return { detected: false, signals: [], terms: effective };
  }

  const stripped = stripFencedCode(text.replace(/\r\n/g, '\n')).text;
  if (stripped.trim().length === 0) {
    return { detected: false, signals: [], terms: effective };
  }

  const signals: ApiCoverageSignal[] = [];
  const seen = new Set<string>();
  const lines = stripped.split('\n');

  const hasCompoundTerms = effective.verbs.length > 0 && effective.nouns.length > 0;
  // Trailing boundary is a LOOKAHEAD (not consumed) so back-to-back terms
  // separated by one boundary char are both found.
  const verbRe = hasCompoundTerms
    ? new RegExp(
        '(^|[^a-zA-Z0-9])(' + effective.verbs.map(escapeRegex).join('|') + ')(?=[^a-zA-Z0-9]|$)',
        'gi',
      )
    : null;
  const nounRe = hasCompoundTerms
    ? new RegExp(
        '(^|[^a-zA-Z0-9])(' + effective.nouns.map(escapeRegex).join('|') + ')(?=[^a-zA-Z0-9]|$)',
        'gi',
      )
    : null;
  const surfaceRe = new RegExp(SERVICE_SURFACE_API_RE.source, 'g');

  const nounSet = new Set(effective.nouns);

  const emitPair = (vTerm: string, nTerm: string, snippetLine: string): void => {
    const key = `${vTerm}+${nTerm}`;
    if (seen.has(key)) return;
    seen.add(key);
    signals.push({ verb: vTerm, noun: nTerm, snippet: makeSnippet(snippetLine, nTerm) });
  };

  for (const rawLine of lines) {
    // Inline code spans are code, not prose — mask them (length-preserving so
    // offsets keep lining up), but keep package-shaped span content as noun
    // evidence (#2365 review FN-4: `stripe-sdk` names a dependency).
    const inlineSpans = scanInlineCodeSpans(rawLine);
    let line = rawLine;
    const spanNouns: TermMatch[] = [];
    for (const s of inlineSpans) {
      line = line.slice(0, s.start) + ' '.repeat(s.end - s.start) + line.slice(s.end);
      const content = s.content.trim();
      if (content.length === 0 || /\s/.test(content)) continue;
      const segs = content.toLowerCase().split(/[^a-z0-9]+/).filter(Boolean);
      if (segs.length < 2) continue; // a bare `api` span is a code identifier
      const hit = segs.find((seg) => nounSet.has(seg));
      if (hit) spanNouns.push({ term: hit, start: s.start, end: s.end });
    }

    // Path-shaped tokens (routes, file names, URLs) are references, not prose.
    const { masked, urlNouns } = scanLineTokens(line, nounRe, nounSet);
    const clauses = splitClauses(masked);
    const extraNouns = urlNouns.concat(spanNouns);

    // (a) compound verb+noun — SAME CLAUSE ONLY. There is no word-gap cap (a cap
    // cannot tell a long genuine clause from a long internal one) and no
    // cross-clause binding (see the note by CLAUSE_BOUNDARY_CHARS): the clause
    // boundary is the whole relationship test. Nouns are NOT filtered on
    // "internal" qualification here — "integrate the internal API" is a
    // fail-closed positive; the declaration dismisses it if wrong.
    //
    // #2784: negation suppression. A clause that pairs an integration verb with
    // an API noun but the verb itself is directly negated (e.g. "does not
    // integrate", "integrates no external API") is suppressed. The check is
    // scoped to the verb's immediate context (the word directly before the
    // verb, or the word directly between verb and noun) — NOT a blanket
    // clause-wide scan, because "without changing runtime dependencies" in a
    // long clause does NOT negate the integration.
    // KNOWN LIMIT (deliberate, not a bug to fix later): a negation further than
    // 2 words before the verb, or a clause where the noun precedes the verb, is
    // NOT suppressed — e.g. "Ships without any API integration." is NOT
    // suppressed today (pinned by a test in tests/api-coverage.test.cjs).
    // detectApiIntegration is fail-closed by design: an unsuppressed false
    // positive costs a one-line COVERAGE.md declaration, while widening the
    // window trades that for a silent false negative on a blocking gate.
    if (verbRe && nounRe) {
      for (const clause of clauses) {
        const verbs = collectTermMatches(verbRe, clause.text);
        if (verbs.length === 0) continue;
        // #2784: check if any verb is immediately preceded by a negation
        // qualifier (within 2 words before the verb match).
        //
        // OFFSET NOTE: `v.start`/`n.start` (from collectTermMatches below and
        // above) are already CLAUSE-LOCAL — collectTermMatches was called with
        // `clause.text`, not the full line — and so is `clauseText`
        // (`clause.text.toLowerCase()`). They must be used AS-IS to index into
        // `clauseText`; do not re-base them against `clause.start` (that field
        // is the clause's offset within the LINE, a different coordinate space,
        // used only to map line-level spans like `extraNouns`/`masked` into a
        // clause). Subtracting `clause.start` here double-offsets the slice
        // bounds for every clause after the first on a line (#3127 follow-up).
        const clauseText = clause.text.toLowerCase();
        const hasNegatedVerb = verbs.some((v) => {
          const before = clauseText.slice(Math.max(0, v.start - 20), v.start);
          const beforeWords = before.split(/\s+/).filter(Boolean).slice(-2);
          return beforeWords.some((w: string) => NEGATION_QUALIFIERS.has(w.replace(/[^a-z']/g, '')));
        });
        // Also check if "no"/"zero"/"none" appears between the verb and the noun.
        const nouns = collectTermMatches(nounRe, clause.text);
        const nounTerms = new Set(nouns.map((t) => t.term));
        for (const u of extraNouns) {
          if (u.start >= clause.start && u.end <= clause.start + clause.text.length) {
            nounTerms.add(u.term);
          }
        }
        if (nounTerms.size === 0) continue;
        // Check for negation between verb and noun.
        //
        // #3127 regression: the original form of this check was
        // O(verbs × nouns), re-slicing and re-splitting the clause text for
        // every (verb, noun) pair — effectively cubic in clause length (a
        // clause of N repeated "integrate api" pairs did O(N^2) pair checks,
        // each doing an O(N) slice/split). On a clause with 800 repeated
        // pairs this took ~8.5s; fast-check's property test then generated
        // documents large enough to hang the whole test file past node:test's
        // 600s timeout. It ALSO subtracted `clause.start` from `v.start`/
        // `n.start` before slicing `clauseText` — but `v.start`/`n.start` are
        // already local to `clause.text` (collectTermMatches was called with
        // clause.text, not the full line), and `clauseText` is exactly
        // `clause.text.toLowerCase()`. So that subtraction double-offset the
        // slice bounds for every clause after the first on a line, sliding
        // (and for negative results, JS's negative-index slice() wraparound
        // non-monotonically re-mapping) the window to characters unrelated to
        // the verb/noun pair — an independent latent bug, fixed here as part
        // of establishing a well-defined O(1) predicate (a piecewise/clamped
        // window has no single "widest span" to reason about at all).
        //
        // EXACT-EQUIVALENCE, single pass: the predicate is "does any pair
        // (v, n) with n.start > v.start have a negation token in the span
        // (v.end, n.start)". Every such span is a SUBSET of the widest
        // possible span for a given noun: [min(v.end) over verbs valid for
        // that noun, n.start). And since that window only widens as a
        // noun's start increases (more verbs become valid, and the noun
        // bound itself grows), the single widest span across the WHOLE
        // clause is anchored at the noun with the maximum start, using the
        // minimum verb-end among verbs valid for THAT noun (not the global
        // minimum verb-end, which could belong to a verb that starts after
        // this noun and so is never a valid pairing with it — a mismatch
        // that would either miss or falsely include a negation). If that one
        // substring contains no negation token, no narrower pair-specific
        // substring can either; if it does, the (minVerb, maxNoun) pair
        // itself contains it. This drops the check to O(verbs + nouns).
        let hasNegatedNoun = false;
        if (nouns.length > 0) {
          let minVerbStart = Infinity;
          for (const v of verbs) if (v.start < minVerbStart) minVerbStart = v.start;
          let maxNounStart = -Infinity;
          for (const n of nouns) if (n.start > maxNounStart) maxNounStart = n.start;
          if (maxNounStart > minVerbStart) {
            let minQualifyingVerbEnd = Infinity;
            for (const v of verbs) {
              if (v.start < maxNounStart) {
                const vEnd = v.start + v.term.length;
                if (vEnd < minQualifyingVerbEnd) minQualifyingVerbEnd = vEnd;
              }
            }
            const between = clauseText.slice(minQualifyingVerbEnd, maxNounStart);
            const betweenWords = between.split(/\s+/).filter(Boolean);
            hasNegatedNoun = betweenWords.some((w: string) =>
              NEGATION_NOUN_TOKENS.has(w.replace(/[^a-z']/g, '')),
            );
          }
        }
        if (hasNegatedVerb || hasNegatedNoun) continue;
        for (const vTerm of new Set(verbs.map((t) => t.term))) {
          for (const nTerm of nounTerms) emitPair(vTerm, nTerm, rawLine);
        }
      }
    }

    // (b) explicit <Service> API|SDK|REST|GraphQL surface — scan every candidate
    // in every clause (a rejected first candidate must not shadow a later
    // genuine service; #2365 review C-1).
    for (const clause of clauses) {
      surfaceRe.lastIndex = 0;
      let m: RegExpExecArray | null;
      while ((m = surfaceRe.exec(clause.text)) !== null) {
        const svc = m[1] || '';
        const svcLower = svc.toLowerCase();
        // Reject capitalized sentence starters ("The API"), locality/protocol
        // descriptors ("Internal API", "REST API"), compound modifiers
        // ("Resolver-only API"), and services qualified first-party
        // ("internal Payments API"). A real vendor name is none of these.
        if (SERVICE_STOPWORDS.has(svcLower)) continue;
        if (SURFACE_DESCRIPTOR_WORDS.has(svcLower)) continue;
        if (COMPOUND_MODIFIER_RE.test(svc)) continue;
        if (isInternallyQualified(masked, clause.start + m.index)) continue;
        const noun = (m[2] || '').toLowerCase();
        const key = `surface+${noun}`;
        if (seen.has(key)) continue;
        seen.add(key);
        signals.push({ verb: '(surface)', noun, snippet: makeSnippet(rawLine, svc) });
      }
    }
  }

  return { detected: signals.length > 0, signals, terms: effective };
}


/** True when the word IMMEDIATELY ADJACENT before `offset` is a locality
 *  descriptor ("internal Payments API") — first-party qualification is negative
 *  evidence for an EXTERNAL-API signal. Only plain spaces/tabs may separate the
 *  descriptor from the service: any intervening punctuation means the descriptor
 *  belongs to a prior clause/sentence and must NOT qualify ("The cache is
 *  private. Stripe API …" — `private` is a different sentence; #2365 review).
 *  Looks back through a BOUNDED window, not the whole prefix, to stay linear. */
const QUALIFIER_LOOKBACK = 24; // longest descriptor ("first-party") + separators
function isInternallyQualified(masked: string, offset: number): boolean {
  const from = offset > QUALIFIER_LOOKBACK ? offset - QUALIFIER_LOOKBACK : 0;
  const window = masked.slice(from, offset);
  // Only whitespace and markdown emphasis/wrapper markers (`*_~\`) may separate
  // the descriptor from the service, so "The **internal** Payments API" still
  // qualifies — but NOT a clause/sentence boundary, so "…is private. Stripe API"
  // does not (the descriptor is a different sentence; #2365 review).
  const m = /([A-Za-z0-9'-]+)[\s*_~`]*$/.exec(window);
  if (!m) return false;
  // A word truncated by the window start is not a descriptor match (its real
  // start lies before the window) — fail toward detection.
  if (from > 0 && m.index === 0 && /[A-Za-z0-9'-]/.test(masked[from - 1])) return false;
  return INTERNAL_DESCRIPTORS.has(m[1].toLowerCase());
}

// ─── Coverage matrix parse / validate / render ────────────────────────────────

export type CoverageDecision = 'INTEGRATE' | 'OPT-OUT';

export interface CoverageRow {
  capability: string;
  decision: CoverageDecision;
  reason: string;
}

/** #2365 acceptance #5: a first-class "this phase integrates no external API"
 *  declaration — the legitimate alternative to fabricating a matrix row for a
 *  capability that does not exist. Like an OPT-OUT row, it must carry a
 *  reason: the declaration is a reasoned decision, not a bypass. */
export interface CoverageNoneDeclaration {
  none: true;
  reason: string;
}

export interface CoverageParseResult {
  rows: CoverageRow[];
  errors: string[];
  format: 'table' | 'json' | 'none';
  declaration: CoverageNoneDeclaration | null;
}

export interface CoverageValidationResult {
  valid: boolean;
  errors: string[];
  counts: { surface: number; integrate: number; optout: number };
  /** True when a valid no-integration declaration (and no rows) satisfied the gate. */
  none_declared?: boolean;
}

/** Matches a declaration line such as
 *  `No external API integration: <reason>` (also `**bold**` and em-dash
 *  separators). The reason is REQUIRED — a bare declaration does not parse.
 *  Deliberately NOT matched: blockquoted lines (`> No external …` is quoted
 *  text, not a declaration) and anything inside fenced code or HTML comments
 *  (both stripped before the scan; #2365 review C-3). */
const NO_INTEGRATION_DECLARATION_RE =
  /^\s*(?:\*\*)?no external api integration(?:\*\*)?\s*(?:[:—–-]|--)\s*(\S[^\n]*)$/im;
const HTML_COMMENT_RE = /<!--[\s\S]*?-->/g;

const VALID_DECISIONS = new Set<CoverageDecision>(['INTEGRATE', 'OPT-OUT']);

/**
 * Parse a coverage matrix from COVERAGE.md. Accepts two bijective formats:
 *
 *  1. Markdown table (canonical, human-editable):
 *       | capability | decision | reason |
 *       |---|---|---|
 *       | search | INTEGRATE | |
 *       | playlists | OPT-OUT | not needed yet |
 *
 *  2. Fenced ```coverage JSON block (machine-generated):
 *       ```coverage
 *       [ {"capability":"search","decision":"INTEGRATE","reason":""}, ... ]
 *       ```
 *
 * Rows are trimmed; decisions upper-cased; missing reason → "". Returns
 * `{ rows: [], errors: [], format: 'none' }` for empty/non-matrix input.
 */
export function parseCoverageMatrix(text: unknown): CoverageParseResult {
  const out: CoverageParseResult = { rows: [], errors: [], format: 'none', declaration: null };
  if (typeof text !== 'string') return out;
  const src = text.replace(/\r\n/g, '\n');

  // #2365 acceptance #5: a "no external API integration" declaration. Scanned
  // on fence-stripped, comment-stripped text so an example inside a code block
  // or an HTML comment does not count.
  const declMatch = NO_INTEGRATION_DECLARATION_RE.exec(
    stripFencedCode(src).text.replace(HTML_COMMENT_RE, ''),
  );
  if (declMatch) {
    out.declaration = { none: true, reason: (declMatch[1] || '').trim() };
  }

  // (1) fenced ```coverage JSON block takes precedence if present.
  // Case-insensitive info string (```coverage and ```Coverage are both legal CommonMark).
  const fenceBody = extractFencedBlock(src, 'coverage');
  if (fenceBody) {
    out.format = 'json';
    let parsed: unknown;
    try {
      parsed = JSON.parse(fenceBody);
    } catch {
      out.errors.push('fenced ```coverage block is not valid JSON');
      return out;
    }
    if (!Array.isArray(parsed)) {
      out.errors.push('fenced ```coverage block must be a JSON array');
      return out;
    }
    for (let i = 0; i < parsed.length; i++) {
      const row = rowFromJson(parsed[i]);
      if ('error' in row) {
        out.errors.push(`row[${i}]: ${row.error}`);
        continue;
      }
      out.rows.push(row);
    }
    return out;
  }

  // (2) markdown table — collect rows from coverage matrix tables only (#2366).
  // Track whether we are inside a recognized coverage matrix (after a header
  // row, before a non-pipe line ends the table). This prevents summary tables
  // elsewhere in the file from being parsed as data (#2366 bug 1) and allows
  // multi-section matrices with repeated headers (#2366 bug 2).
  const lines = src.split('\n');
  let inMatrix = false;
  for (const line of lines) {
    const trimmed = line.trim();
    if (!trimmed.startsWith('|')) {
      inMatrix = false;
      continue;
    }
    const cells = trimmed.slice(1, trimmed.endsWith('|') ? -1 : trimmed.length).split('|');
    if (cells.length < 2) continue;
    const cleaned = cells.map((c) => c.trim());
    // skip separator rows (|---|---|); require ≥3 dashes so a literal "-" cell
    // is not mistaken for a separator.
    if (cleaned.every((c) => /^:?-{3,}:?$/.test(c))) continue;
    // Strip markdown emphasis (**, *, __, _, `) from the decision cell before
    // comparison so **OPT-OUT** parses correctly (#2366 bug 3).
    const decisionCell = (cleaned[1] || '').replace(/[*_`]/g, '').trim().toUpperCase();
    // header detection — recognized by 'capability' in column 0; allows multiple
    // headers for multi-section matrices (#2366 bug 2).
    if (cleaned[0].toLowerCase() === 'capability') {
      inMatrix = true;
      if (out.format === 'none') out.format = 'table';
      continue;
    }
    // Only parse data rows from inside a recognized coverage matrix table.
    // A pipe-table outside the matrix (e.g., a summary table) is ignored (#2366 bug 1).
    if (!inMatrix) continue;
    if (!VALID_DECISIONS.has(decisionCell as CoverageDecision)) {
      // A row that otherwise looks like data (≥3 cells, non-empty capability)
      // but carries a malformed decision is a real error, not a row to skip
      // silently — otherwise a single typo'd row collapses the matrix to
      // "empty" and the user sees a confusing message.
      if (cleaned.length >= 3 && cleaned[0]) {
        out.errors.push(`row: decision "${decisionCell}" not in {INTEGRATE, OPT-OUT}`);
      }
      continue;
    }
    if (out.format === 'none') out.format = 'table';
    // A coverage row has exactly 3 cells. Extra cells mean an unescaped pipe in
    // a value silently corrupted the row — surface it rather than parse garbage.
    if (cleaned.length > 3) {
      out.errors.push(`row: ${cleaned.length} columns (expected 3 — unescaped pipe in a cell?)`);
    }
    out.rows.push({
      capability: cleaned[0] || '',
      decision: decisionCell as CoverageDecision,
      reason: (cleaned[2] ?? '').trim(),
    });
  }
  return out;
}

function rowFromJson(v: unknown): CoverageRow | { error: string } {
  if (!v || typeof v !== 'object' || Array.isArray(v)) return { error: 'not an object' };
  const o = v as Record<string, unknown>;
  const capability = typeof o['capability'] === 'string' ? o['capability'].trim() : '';
  if (!capability) return { error: 'missing/empty "capability"' };
  const dRaw = typeof o['decision'] === 'string' ? o['decision'].trim().toUpperCase() : '';
  if (!VALID_DECISIONS.has(dRaw as CoverageDecision)) {
    return { error: `decision "${dRaw}" not in {INTEGRATE, OPT-OUT}` };
  }
  const reason = typeof o['reason'] === 'string' ? o['reason'].trim() : '';
  return { capability, decision: dRaw as CoverageDecision, reason };
}

/**
 * Validate a parsed matrix. A matrix is valid when:
 *   - it is non-empty (acceptance #1: "enumerating the API surface"),
 *   - every capability name is non-empty,
 *   - every decision is INTEGRATE or OPT-OUT (enforced by parser, re-checked
 *     here for defense-in-depth),
 *   - every OPT-OUT row carries a non-empty reason (acceptance #2).
 *
 * Un-enumerated remainder is not representable in the format — the gate blocks
 * when an integration is detected and NO matrix exists. This validator catches
 * a malformed/partial matrix that does exist.
 */
export function validateCoverageMatrix(text: unknown): CoverageValidationResult {
  const parsed = parseCoverageMatrix(text);
  const errors = [...parsed.errors];
  const rows = parsed.rows;

  // #2365 acceptance #5: a reasoned no-integration declaration with no rows
  // satisfies the gate. A declaration ALONGSIDE rows is contradictory — the
  // file must say one thing.
  if (parsed.declaration) {
    if (rows.length > 0) {
      errors.push(
        'declares "no external API integration" but also contains coverage rows — remove the declaration or the rows',
      );
    } else {
      if (parsed.declaration.reason.length > REASON_MAX_LEN) {
        errors.push(`declaration reason exceeds ${REASON_MAX_LEN} chars`);
      }
      const valid = errors.length === 0;
      return {
        valid,
        errors,
        counts: { surface: 0, integrate: 0, optout: 0 },
        none_declared: valid,
      };
    }
  }

  if (rows.length === 0) {
    if (errors.length === 0) errors.push('matrix is empty — no capabilities enumerated');
    return { valid: false, errors, counts: { surface: 0, integrate: 0, optout: 0 } };
  }

  const seen = new Set<string>();
  for (let i = 0; i < rows.length; i++) {
    const row = rows[i];
    if (!row.capability) {
      errors.push(`row[${i}]: empty capability name`);
    } else {
      // Format contract + prompt-injection bound: cell values must be short,
      // single-line, pipe-free prose (the matrix is a markdown table whose
      // content flows into the gate message). Pipes/newlines would corrupt the
      // table and let a COVERAGE.md inject unbounded text into the seal message.
      if (/[|\n\r]/.test(row.capability)) {
        errors.push(`row[${i}]: capability contains a pipe or newline (unsupported in a table cell)`);
      }
      if (row.capability.length > CAPABILITY_MAX_LEN) {
        errors.push(`row[${i}]: capability exceeds ${CAPABILITY_MAX_LEN} chars`);
      }
    }
    if (row.reason && /[|\n\r]/.test(row.reason)) {
      errors.push(`row[${i}]: reason contains a pipe or newline (unsupported in a table cell)`);
    }
    if (row.reason.length > REASON_MAX_LEN) {
      errors.push(`row[${i}]: reason exceeds ${REASON_MAX_LEN} chars`);
    }
    const key = row.capability.toLowerCase();
    if (key && seen.has(key)) errors.push(`row[${i}]: duplicate capability`);
    if (key) seen.add(key);
    if (!VALID_DECISIONS.has(row.decision)) {
      errors.push(`row[${i}]: decision not in {INTEGRATE, OPT-OUT}`);
    }
    if (row.decision === 'OPT-OUT' && !row.reason) {
      errors.push(`row[${i}]: OPT-OUT missing reason`);
    }
  }

  const counts = {
    surface: rows.length,
    integrate: rows.filter((r) => r.decision === 'INTEGRATE').length,
    optout: rows.filter((r) => r.decision === 'OPT-OUT').length,
  };

  return { valid: errors.length === 0, errors, counts };
}

/** Render rows back to the canonical markdown-table format (bijective with parse). */
export function renderCoverageMatrix(rows: readonly CoverageRow[]): string {
  const body = rows
    .map((r) => `| ${r.capability} | ${r.decision} | ${r.reason} |`)
    .join('\n');
  return `| capability | decision | reason |\n|---|---|---|\n${body}`;
}

// ── CLI entry point ──────────────────────────────────────────────────────────
// Reads phase-scope text from STDIN (not argv) to avoid OS ARG_MAX limits.
// Invoked by workflow bash as: echo "$SCOPE" | node .../api-coverage.cjs [--json]
// Exit 0 = integration detected, 1 = none, 2 = startup error. Mirrors
// assumption-delta.cjs / ui-safety-gate.cjs.

if (require.main === module) {
  const argv = process.argv.slice(2);
  const wantJson = argv.includes('--json');

  let termsOverride: Partial<ApiCoverageTermSet> | undefined;
  const verbsIdx = argv.indexOf('--verbs');
  const verbsVal = verbsIdx !== -1 ? argv[verbsIdx + 1] : undefined;
  const nounsIdx = argv.indexOf('--nouns');
  const nounsVal = nounsIdx !== -1 ? argv[nounsIdx + 1] : undefined;
  // A non-empty, non-flag value is an override. An EMPTY value ("") restores
  // the curated defaults (does NOT silently zero the vocabulary).
  const verbsOverride = typeof verbsVal === 'string' && verbsVal.length > 0 && !verbsVal.startsWith('-');
  const nounsOverride = typeof nounsVal === 'string' && nounsVal.length > 0 && !nounsVal.startsWith('-');
  if (verbsOverride || nounsOverride) {
    termsOverride = {};
    if (verbsOverride) {
      termsOverride.verbs = verbsVal.split(',').map((t) => t.trim().toLowerCase()).filter(Boolean);
    }
    if (nounsOverride) {
      termsOverride.nouns = nounsVal.split(',').map((t) => t.trim().toLowerCase()).filter(Boolean);
    }
  }

  const chunks: string[] = [];
  process.stdin.setEncoding('utf-8');
  process.stdin.on('data', (chunk: string) => chunks.push(chunk));
  process.stdin.on('end', () => {
    const input = chunks.join('');
    const result = detectApiIntegration(input, termsOverride);
    if (wantJson) {
      process.stdout.write(JSON.stringify(result) + '\n');
    }
    process.exit(result.detected ? 0 : 1);
  });
  process.stdin.on('error', (err: Error) => {
    process.stderr.write(`ERROR: api-coverage.cjs stdin read failed: ${err.message}\n`);
    process.exit(2);
  });
}