@pipeworx/mcp-meps 0.1.0

package/src/index.ts

@@ -0,0 +1,998 @@
+interface McpToolDefinition {
+  name: string;
+  description: string;
+  /** Human-facing one-liner (fleet #1967). Optional; consumers fall back to
+   *  description. Kept in step with shared/src/types.ts — scripts/lib/
+   *  check-inlined-types.mjs reports drift at publish time. */
+  summary?: string;
+  inputSchema: {
+    type: 'object';
+    properties: Record<string, unknown>;
+    required?: string[];
+    anyOf?: Array<{ required: string[] }>;
+    oneOf?: Array<{ required: string[] }>;
+    allOf?: Array<{ required: string[] }>;
+  };
+  outputSchema?: Record<string, unknown>;
+}
+
+interface McpToolExport {
+  tools: McpToolDefinition[];
+  callTool: (name: string, args: Record<string, unknown>) => Promise<unknown>;
+  meter?: { credits: number };
+  cost?: Record<string, unknown>;
+  provider?: string;
+}
+
+/**
+ * One place to turn a failed `fetch` into an error a caller can act on.
+ *
+ * Nearly every pack was written the same way:
+ *
+ *     if (!res.ok) throw new Error(`Unsplash: ${res.status}`);
+ *
+ * which discards the response body — and the body is usually where the upstream
+ * says what was actually wrong ("**symbol** not found: GBP", "parameter `year`
+ * out of range", "unknown taxonomy id"). The caller gets a number, cannot
+ * self-correct, and retries the same broken call. A 2026-07-31 sweep found this
+ * shape in 481 of 1,400 packs, 47 of them PLATFORM-keyed.
+ *
+ * It also hides bugs one level down. Two of the first three packs audited had a
+ * second defect that only existed because of this line: unsplash's rate-limit
+ * branch sat BELOW a catch-all and was unreachable, and bea-gov parsed
+ * `BEAAPI.Error.APIErrorDescription` below a `!res.ok` throw that made the
+ * parsing dead code for every non-200.
+ *
+ * DELIBERATELY NOT A CLASSIFIER. It does not add `user_error:` /
+ * `upstream_down:` prefixes. Those decide which tier a failure lands in, and the
+ * `error` tier is what the daily problem-tools list is built from — it means
+ * "Pipeworx has a defect". A 400 is genuinely ambiguous: often a caller's bad
+ * argument, but sometimes a query WE built wrong (ted-eu comma-joined its CPV
+ * values into something TED rejected, and that bug was found only because it sat
+ * in `error`). Blanket-classifying 400s as caller mistakes would have hidden it.
+ * A pack that KNOWS which it is should keep saying so explicitly; this helper is
+ * for the 481 that say nothing at all.
+ */
+
+/** Longest upstream explanation we'll pass through. Enough for a real message,
+ *  short enough that an HTML page or a stack trace can't swamp the error. */
+
+const MAX_DETAIL = 300;
+
+/**
+ * Default bound for `fetchWithTimeout` when a pack doesn't state its own.
+ *
+ * 25s mirrors the number `epo-ops` landed on after measuring the real failure:
+ * a degraded upstream that doesn't error, it just never answers, and a Worker
+ * sits in `await fetch()` until ITS OWN execution budget kills the request —
+ * which can take minutes, not seconds (epo_ops_search_patents measured 4-8
+ * MINUTE hangs before this existed). 25s is short enough that a caller gets a
+ * fast, actionable error instead of holding the connection, and long enough
+ * that it doesn't false-trip on a merely-slow-but-alive upstream.
+ */
+const DEFAULT_FETCH_TIMEOUT_MS = 25_000;
+
+/**
+ * Read the body of a failed response and fold it into a throwable Error.
+ *
+ * Usage — note the `await`, which is the one thing that makes this a mechanical
+ * change rather than a drop-in:
+ *
+ *     if (!res.ok) throw await httpError(res, 'Unsplash');
+ *
+ * Safe to call on any non-ok response: a body that is missing, empty, unreadable
+ * or HTML degrades to exactly the old `Name: 404` string rather than throwing
+ * something new from inside the error path.
+ */
+async function httpError(res: Response, name: string): Promise<Error> {
+  return new Error(await httpErrorMessage(res, name));
+}
+
+/** The message text without constructing an Error — for packs that need to wrap
+ *  it in their own envelope or add an explicit classification prefix. */
+async function httpErrorMessage(res: Response, name: string): Promise<string> {
+  // The one place a 5xx from a host WE run gets stamped as ours. `res.url` is
+  // the URL the fetch actually resolved to (after redirects), so this is a fact
+  // about the call rather than a guess from the `name` the pack passed in —
+  // reword that label freely, the class does not move. See
+  // internal-host-class.ts; no-op for every third-party upstream, which is why
+  // this touches 481 packs' error text and changes none of it.
+  return markInternalOrigin(
+    `${name}: ${res.status}${detailSuffix(await readDetail(res))}`,
+    res.url,
+    res.status,
+  );
+}
+
+/**
+ * Just the upstream's own explanation — no name, no status.
+ *
+ * For a pack that has already said both in its own sentence. epo-ops reads
+ * `EPO rejected this search as too large (HTTP 413) — ${httpErrorMessage(…)}`,
+ * which rendered as `… (HTTP 413) — EPO: 413.` once the XML detail was being
+ * dropped: the upstream named twice, the status twice, and the one thing EPO
+ * actually said ("Not enough characters before truncation character") nowhere
+ * (fleet #712). Returns '' when the body carries nothing readable, so a caller
+ * can fall back to its own wording.
+ */
+async function upstreamDetail(res: Response): Promise<string> {
+  return readDetail(res);
+}
+
+/**
+ * Read a SUCCESSFUL response as JSON, failing loudly when it isn't JSON.
+ *
+ * `httpError` above only ever runs on `!res.ok`, which leaves the nastier half
+ * of the problem unhandled: an upstream that answers **HTTP 200 with an HTML
+ * page**. A bot wall, a login redirect, a maintenance interstitial and a CDN
+ * error page are all 200s, so `res.ok` is true, and `res.json()` then throws
+ * `Unexpected token '<', "<!DOCTYPE "... is not valid JSON`.
+ *
+ * That string is the problem. It names no upstream, carries no status, and
+ * reads like a parser bug in Pipeworx — so it lands in the `error` tier, which
+ * means "we have a defect", and the caller is told nothing they can act on.
+ * data.govt.nz sat dead behind an Imperva challenge this way and every
+ * status-code health check we own reported it green (7889a845). A zero-length
+ * body has the same shape: `Unexpected end of JSON input`, seen this week on
+ * uk-gazette (83% of external calls) and census.
+ *
+ * UNLIKE `httpError`, this one DOES classify, and the asymmetry is deliberate.
+ * A 400 is genuinely ambiguous — often the caller's bad argument, sometimes a
+ * query we built wrong — so blanket-classifying it would hide our own bugs.
+ * There is no such ambiguity here: **no argument a caller can pass makes a JSON
+ * API return an HTML page.** It is always the upstream, so `upstream_down:` is
+ * a statement of fact rather than a guess, and it keeps these out of the
+ * problem-tools list where they crowd out real defects.
+ *
+ *     const data = await parseJson<Feed>(res, 'UK Gazette');
+ *
+ * Call it only after the `!res.ok` check — on a failed response you want
+ * `httpError`, which mines the body for the upstream's own explanation.
+ */
+async function parseJson<T>(res: Response, name: string): Promise<T> {
+  let raw: string;
+  try {
+    raw = await res.text();
+  } catch {
+    throw new Error(
+      `upstream_down: ${name} returned a body that could not be read (HTTP ${res.status}). ` +
+        'The connection most likely dropped mid-response; retrying is reasonable.',
+    );
+  }
+
+  const type = res.headers.get('content-type') ?? 'no content-type';
+
+  if (!raw.trim()) {
+    throw new Error(
+      `upstream_down: ${name} answered HTTP ${res.status} with an EMPTY body where JSON was expected (${type}). ` +
+        'Nothing about the request can cause this — it is an upstream fault, and the same call may well work on retry.',
+    );
+  }
+
+  // Checked before parsing rather than in the catch, because knowing it is
+  // markup is what turns "we failed to parse something" into "they served a
+  // web page" — the second is diagnosable, the first is not.
+  const head = raw.slice(0, 200).trimStart().toLowerCase();
+  if (head.startsWith('<!doctype') || head.startsWith('<html') || head.startsWith('<?xml')) {
+    const kind = head.startsWith('<?xml') ? 'an XML document' : 'an HTML page';
+    // The summary, not the source. Pasting the first 120 characters of a web
+    // page handed the agent `<!DOCTYPE html><html lang="en"…` — the same leak
+    // this branch exists to describe (fleet #712).
+    throw new Error(
+      `upstream_down: ${name} answered HTTP ${res.status} with ${kind} instead of JSON (${type}). ` +
+        'That is typically a bot wall, a login redirect or a maintenance page — it is returned as a SUCCESS, ' +
+        `so status-code health checks read it as fine. No argument change will get past it. ` +
+        `The page says: ${summarizeErrorBody(raw) || 'nothing readable'}`,
+    );
+  }
+
+  try {
+    return JSON.parse(raw) as T;
+  } catch {
+    throw new Error(
+      `upstream_down: ${name} answered HTTP ${res.status} with a body that is not valid JSON (${type}). ` +
+        `It begins: ${stripMarkup(raw).slice(0, 120) || '(unreadable)'}`,
+    );
+  }
+}
+
+/**
+ * `fetch`, but bounded — the fix for a systemic gap found 2026-08-30: a grep
+ * audit of every pack's `mcps/*\/src/index.ts` found 1,339 of ~1,500 call
+ * `fetch()` with NO timeout guard anywhere in the file. Two of those
+ * (epo-ops, statcan) were confirmed live-hanging for 4-8 minutes before this
+ * existed — every unguarded call carries the same risk, just unconfirmed.
+ *
+ * Mirrors the `epoFetch` wrapper `mcps/epo-ops/src/index.ts` shipped first:
+ * bound the request with `AbortSignal.timeout`, and on a timeout/abort throw
+ * an `upstream_down:` error that names the upstream and the bound rather than
+ * letting the raw `TimeoutError`/`AbortError` (which names neither) propagate.
+ * `upstream_down:` is deliberate, same reasoning as `parseJson` above — no
+ * argument a caller passes can make an upstream hang, so it is always the
+ * upstream's fault, and marking it that way keeps a slow API off the
+ * problem-tools list where it would crowd out our own defects.
+ *
+ * Usage — a mechanical swap for a bare `fetch(url, init)`:
+ *
+ *     const res = await fetchWithTimeout(url, init, 'Some API');
+ *
+ * Pass `timeoutMs` as a fourth argument to override the default for a pack
+ * with a known-slower upstream; the label should be the same short name you'd
+ * pass to `httpError`/`httpErrorMessage` for that call.
+ */
+async function fetchWithTimeout(
+  url: string | URL,
+  init: RequestInit = {},
+  name: string,
+  timeoutMs: number = DEFAULT_FETCH_TIMEOUT_MS,
+): Promise<Response> {
+  try {
+    return await fetch(url, { ...init, signal: AbortSignal.timeout(timeoutMs) });
+  } catch (err) {
+    if (err instanceof Error && (err.name === 'TimeoutError' || err.name === 'AbortError')) {
+      // States the OBSERVATION (no response in N seconds), not a diagnosis.
+      // "appears to be degraded" is an inference about the vendor that we have
+      // not checked, and it is wrong in a way that misdirects whoever reads it:
+      // a timeout from a Worker can equally mean OUR egress is blocked.
+      //
+      // Measured today (2026-09-01, fleet #1047): every call to
+      // mainnet.base.org failed from the x402 facilitator while the identical
+      // request from a laptop returned 200. Base was entirely healthy; the
+      // public RPC refuses Cloudflare Worker egress. Had this message fired
+      // there it would have blamed Base by name, and the next person would have
+      // waited for a vendor outage to clear that did not exist.
+      // A timeout has no status to test — there is no response at all — so
+      // `markInternalOrigin` is called without one: an origin we run that never
+      // answered is an availability failure by definition. This is the half of
+      // fleet #1096 with neither a SQLSTATE nor a status code to key on.
+      throw new Error(
+        markInternalOrigin(
+          `upstream_down: ${name} did not respond within ${timeoutMs / 1000}s. ` +
+            `That can be ${name} being slow or down, or this environment being unable to reach it ` +
+            `(some hosts refuse datacenter/Worker egress) — retry shortly, and check reachability ` +
+            `from elsewhere before concluding ${name} is down.`,
+          url,
+        ),
+      );
+    }
+    throw err;
+  }
+}
+
+function detailSuffix(detail: string): string {
+  return detail ? ` — ${detail}` : '';
+}
+
+async function readDetail(res: Response): Promise<string> {
+  let raw: string;
+  try {
+    raw = await res.text();
+  } catch {
+    // Body already consumed, or the connection died mid-read. The status alone
+    // is still worth throwing — never let the error path throw its own error.
+    return '';
+  }
+  return summarizeErrorBody(raw);
+}
+
+/**
+ * Turn ANY error body — JSON, HTML, XML or plain text — into one short phrase
+ * that never contains markup.
+ *
+ * This used to just drop an HTML or XML body on the floor, on the reasoning
+ * that markup crowds out the status. That was half right. Dropping it loses the
+ * one sentence a caller could have acted on: an `Access Denied` title, an SDMX
+ * `<message:Error>` text, an OPS fault string. A 2026-08-30 support sweep
+ * measured 13 of 291 caller-facing error rows carrying a raw page or document
+ * verbatim, across 11 packs, and in every one of them the useful content —
+ * "Access Denied", "Invalid country code", "SCRAPE_TIMEOUT" — was in there,
+ * buried in markup the agent had to parse out of a string (fleet #712).
+ *
+ * So: extract the meaning, discard the markup. The output is passed through
+ * `stripMarkup` unconditionally, which is what lets `check:error-body-leak`
+ * assert mechanically that no caller-facing message can contain `<?xml`,
+ * `<!DOCTYPE` or `<html`.
+ */
+function summarizeErrorBody(raw: string): string {
+  if (!raw || !raw.trim()) return '';
+
+  const head = raw.slice(0, 400).trimStart().toLowerCase();
+
+  // An HTML error page (Cloudflare interstitial, nginx default, a login
+  // redirect) says what it is in its <title>, and almost nowhere else.
+  if (head.startsWith('<!doctype') || head.startsWith('<html')) {
+    const title = htmlTitle(raw);
+    return title
+      ? `${title} (upstream returned an HTML error page, not an API response)`
+      : 'upstream returned an HTML error page, not an API response';
+  }
+
+  // XML fault documents — EPO OPS, SDMX (`<message:Error>`), SOAP faults. The
+  // human sentence sits in a child element whose tag name says what it is.
+  if (head.startsWith('<?xml') || head.startsWith('<')) {
+    const fault = xmlFaultText(raw);
+    return fault
+      ? `${stripMarkup(fault).slice(0, MAX_DETAIL)} (from the upstream's XML error document)`
+      : 'upstream returned an XML error document with no readable message';
+  }
+
+  // Most JSON error bodies bury one human sentence among ids and echoed request
+  // params. Prefer that sentence; fall back to the whole body when the shape is
+  // unfamiliar, since an unfamiliar shape is exactly when we can least afford to
+  // guess wrong and show nothing.
+  const fromJson = messageFromJson(raw);
+  return stripMarkup(fromJson ?? raw).slice(0, MAX_DETAIL);
+}
+
+/** The `<title>` of an HTML error page, or its first `<h1>` — the two places a
+ *  bot wall, a 502 and an "Access Denied" all state what happened. */
+function htmlTitle(raw: string): string | null {
+  const head = raw.slice(0, 4000);
+  for (const re of [/<title[^>]*>([\s\S]*?)<\/title>/i, /<h1[^>]*>([\s\S]*?)<\/h1>/i]) {
+    const m = re.exec(head);
+    const text = m ? stripMarkup(m[1]) : '';
+    if (text) return text.slice(0, 160);
+  }
+  return null;
+}
+
+/** Tag names that carry the explanation in an XML fault document, namespace
+ *  prefix optional (`<message:Error>`, `<com:Text>`, `<faultstring>`). */
+const XML_FAULT_TAG_RE =
+  /<(?:[A-Za-z0-9_.-]+:)?(?:text|message|description|faultstring|reason|detail|title|errormessage|error)\b[^>]*>([^<]{2,400})</i;
+
+function xmlFaultText(raw: string): string | null {
+  const head = raw.slice(0, 8000);
+  const tagged = XML_FAULT_TAG_RE.exec(head);
+  if (tagged && tagged[1].trim()) return tagged[1];
+
+  // Nothing conventionally named — take the longest text node instead. A fault
+  // document with one sentence in an oddly named element is still readable;
+  // returning nothing at all is not.
+  let best = '';
+  for (const m of head.matchAll(/>([^<>]{8,400})</g)) {
+    const text = m[1].trim();
+    if (text.length > best.length) best = text;
+  }
+  return best || null;
+}
+
+/**
+ * Remove every tag and stray angle bracket, then collapse whitespace.
+ *
+ * Applied to everything on the way out, including the JSON and plain-text
+ * paths, because an upstream is free to embed markup in a JSON string field —
+ * and a leak is a leak regardless of which branch produced it.
+ */
+function stripMarkup(s: string): string {
+  return collapse(decodeEntities(s.replace(/<[^>]*>/g, ' ')).replace(/[<>]/g, ' '));
+}
+
+/** The handful of entities that show up in error-page titles. Decoded AFTER
+ *  tags are stripped and BEFORE the angle-bracket sweep, so `&lt;script&gt;`
+ *  in a title cannot decode into markup that survives — EMBL-EBI's ChEMBL 500
+ *  page renders as `500 Internal Server Error &lt; EMBL-EBI` otherwise. */
+function decodeEntities(s: string): string {
+  return s
+    .replace(/&(?:amp|#0*38);/gi, '&')
+    .replace(/&(?:lt|#0*60);/gi, '<')
+    .replace(/&(?:gt|#0*62);/gi, '>')
+    .replace(/&(?:quot|#0*34);/gi, '"')
+    .replace(/&(?:#0*39|apos|#x0*27);/gi, "'")
+    .replace(/&nbsp;/gi, ' ');
+}
+
+/** The conventional "what went wrong" field, under any of the names upstreams
+ *  actually use. Checked in order; first non-empty string wins. */
+const MESSAGE_KEYS = [
+  'message', 'error_message', 'errorMessage', 'detail', 'details',
+  'description', 'error_description', 'reason', 'title', 'fault',
+];
+
+function messageFromJson(raw: string): string | null {
+  let parsed: unknown;
+  try {
+    parsed = JSON.parse(raw);
+  } catch {
+    return null;
+  }
+  return pickMessage(parsed, 0);
+}
+
+function pickMessage(node: unknown, depth: number): string | null {
+  // Two levels covers `{error: {message}}` and `{errors: [{detail}]}`, the two
+  // shapes that account for nearly all of them, without walking a large payload.
+  if (depth > 2 || node == null) return null;
+
+  if (typeof node === 'string') return node.trim() || null;
+
+  if (Array.isArray(node)) {
+    for (const item of node) {
+      const found = pickMessage(item, depth + 1);
+      if (found) return found;
+    }
+    return null;
+  }
+
+  if (typeof node !== 'object') return null;
+  const obj = node as Record<string, unknown>;
+
+  for (const key of MESSAGE_KEYS) {
+    const v = obj[key];
+    if (typeof v === 'string' && v.trim()) return v.trim();
+  }
+  // `{error: …}` where error is itself an object or a string — the single most
+  // common wrapper, so it is worth descending into by name rather than scanning
+  // every key and risking picking up an echoed request parameter.
+  for (const key of ['error', 'errors', 'fault', 'Error', 'data']) {
+    if (key in obj) {
+      const found = pickMessage(obj[key], depth + 1);
+      if (found) return found;
+    }
+  }
+  return null;
+}
+
+/** Errors are read in a single line of log output; newlines and runs of
+ *  whitespace make a multi-line body unreadable there. */
+function collapse(s: string): string {
+  return s.replace(/\s+/g, ' ').trim();
+}
+
+/**
+ * Was this failure OUR OWN web service? — the other half of `internal-db-class.ts`.
+ *
+ * fleet #1089 pulled failures from our own Postgres out of `upstream_down` by
+ * keying on the SQLSTATE inside PostgREST's four-key error envelope. That
+ * covered the majority and structurally could not cover the rest: the rest
+ * never reach Postgres, so they carry no SQLSTATE. What was left, measured over
+ * the 24h to 2026-09-02T15:00Z (fleet #1096):
+ *
+ *     5  pipeworx-catalog  get_pack_tools     Pipeworx catalog error: 522 — error code: 522
+ *     3  fleet             fleet_list_open …  upstream_down: Fleet task queue did not respond within 25s
+ *
+ * 521/522/523/526 are Cloudflare saying its edge could not reach an ORIGIN, and
+ * in both of those rows the origin is ours — `gateway.pipeworx.io` for the
+ * catalog pack (it self-fetches when the gateway hasn't injected a manifest),
+ * our own Supabase for fleet. There is no third party anywhere in either call.
+ * Same defect as #1089: our own outage filed under `upstream_down`, the one
+ * class that means "the source is unreachable and there is nothing for us to
+ * fix", which is why the problem-tools triage skips it.
+ *
+ * WHY NOT A WORDING RULE. The obvious fix is to match `fleet db error:` and
+ * `Pipeworx catalog error:` in classifyToolError. Each is emitted from exactly
+ * one site today, so it would work today. It would also rot the first time
+ * somebody rewords a label — silently, and in the direction of hiding our own
+ * outage, which is worse than the bug being fixed. Every prose rule in
+ * error-class.ts has needed widening as packs invented new wording (#409/#450/
+ * #584); that history is most of that file's comment budget.
+ *
+ * WHAT THIS KEYS ON INSTEAD: **the host the call actually reached.** A URL's
+ * hostname is a fact about the call, not a guess about its prose. Two
+ * consequences that a pack-level flag could not give us, and the reason the
+ * flag was rejected:
+ *
+ *   - It describes the CALL, not the pack. `govcon-intel` fans out to our own
+ *     Supabase AND to genuine third parties; `court-listener` holds our cache
+ *     in Supabase and fetches courtlistener.com. An `internallyHosted: true` on
+ *     either pack would relabel a real third-party outage as ours — inventing
+ *     work, which is the same class of error in the opposite direction.
+ *   - It covers every future internal pack for free, instead of one declared
+ *     slug at a time.
+ *
+ * WHY IT SURVIVES A REWORD. The marker below is not matched as a literal by two
+ * separate files. `markInternalOrigin()` writes it and `internalHostMetricsClass()`
+ * reads it, both from the single exported `INTERNAL_ORIGIN_MARKER` constant in
+ * this module — so changing the wording changes both sides in the same edit and
+ * cannot desynchronise them. The pack's own label (`fleet db error:`,
+ * `Pipeworx catalog error:`) is not read at all: reword it freely, the class is
+ * unaffected. That is the property `stripClassPrefix` lacked when it drifted
+ * from its own classifier three times and needed a CI gate to hold them
+ * together.
+ *
+ * WHERE THE 5xx TEST LIVES. `markInternalOrigin` is called from the places that
+ * hold the real `Response` — `httpError`/`httpErrorMessage` and the timeout
+ * branch of `fetchWithTimeout` in `shared/src/http.ts` — so "is this an
+ * availability failure" is decided from the actual status code, never re-derived
+ * by scraping a number out of a sentence. A 404 from our own registry for a slug
+ * that does not exist is a caller's bad argument and is deliberately NOT marked.
+ */
+
+/**
+ * OUR OWN web service was unreachable — not an upstream, and never `upstream_down`.
+ *
+ * ONE value, not three, unlike `internal_db_*`. That split existed because a
+ * slow query, an exhausted pool and an unknown SQLSTATE have different owners
+ * and different fixes. Here there is only one story to tell — an origin we run
+ * did not answer the edge — and one owner. A bucket with no distinct owner per
+ * value is decoration; #724 is what happens when a class holds several
+ * situations, and inventing sub-values ahead of a reason to act on them
+ * differently is the same mistake with the sign flipped.
+ *
+ * METRICS ONLY, exactly like PLATFORM_KEY_ERROR_CLASS and the internal_db
+ * values. `classifyToolError` still answers `upstream_down` for the retry and
+ * hint paths, which only care whether retrying or a sibling tool might work —
+ * and it might. Nothing a caller sees or is charged changes here.
+ *
+ * READ SIDE: this value is in BROKEN_TOOL_CLASSES, FAULT_CLASSES and
+ * ALL_ERROR_CLASSES in `workers/registry-api/src/index.ts`. All three, or it
+ * lands on no dashboard — fleet #721 is the warning, where the #719 split
+ * worked on the write side and was invisible for weeks.
+ */
+const INTERNAL_SERVICE_UNREACHABLE_CLASS = 'internal_service_unreachable';
+
+/**
+ * The token that carries "this origin is ours" from the call site to the
+ * classifier.
+ *
+ * Appended to the error message rather than attached to the Error object,
+ * because the object does not survive the trip: 275 packs return `{ error:
+ * string }` instead of throwing, the gateway reads `observedError` as a string,
+ * and the fleet pack rebuilds its error from a captured status + body across a
+ * retry loop. A property on an Error would be dropped by every one of those
+ * paths and the class would work in tests and vanish in production.
+ *
+ * Written as a sentence rather than a sigil because it is going to be read by
+ * whoever gets the error, and "our own service, not a third party" is the
+ * single most useful thing to tell them — fetchWithTimeout's own comment
+ * (fleet #1047) is about exactly this ambiguity, where blaming a healthy vendor
+ * by name sent the next person waiting for an outage that did not exist.
+ */
+const INTERNAL_ORIGIN_MARKER = ' [pipeworx-hosted origin — our own service, not a third party]';
+
+/**
+ * Supabase's data plane for a project is `<ref>.supabase.co`, where the ref is
+ * exactly twenty lowercase letters (ours is `pqauisounztsgdgfkhke`).
+ *
+ * Matching the shape rather than listing the ref keeps this correct when we add
+ * a project — `supabaseEnv` on a pack entry already points some packs at a
+ * second one — while still excluding `status.supabase.co`, which is Supabase's
+ * own status page and emphatically not our database. Verified 2026-09-02 by
+ * `grep -rhoE '[a-z0-9-]+\.supabase\.(co|in)' mcps shared workers scripts`: the
+ * only real project ref anywhere in the tree is ours, the rest are doc
+ * placeholders (`abc`, `xyz`, `example`) which this pattern also excludes. Same
+ * finding internal-db-class.ts relies on for the PostgREST envelope being ours
+ * by construction.
+ */
+const SUPABASE_PROJECT_HOST = /^[a-z]{20}\.supabase\.(co|in)$/;
+
+/**
+ * Is this a host WE run?
+ *
+ * Deliberately NOT including `*.workers.dev`: plenty of third-party APIs are
+ * hosted on workers.dev, so the suffix says where something runs and not who
+ * owns it. Every internal call we actually make goes to a `pipeworx.io`
+ * hostname or to our Supabase project, both of which are ownership facts.
+ *
+ * Returns false on anything unparseable rather than throwing — this runs inside
+ * an error path, and an error path that can itself throw turns a diagnosable
+ * failure into a mystery.
+ */
+function isPipeworxOrigin(url: string | URL | undefined | null): boolean {
+  if (!url) return false;
+  let host: string;
+  try {
+    host = new URL(url instanceof URL ? url.href : url).hostname.toLowerCase();
+  } catch {
+    return false;
+  }
+  if (host === 'pipeworx.io' || host.endsWith('.pipeworx.io')) return true;
+  return SUPABASE_PROJECT_HOST.test(host);
+}
+
+/**
+ * Append the marker when this failure was OUR origin failing to answer.
+ *
+ * `status` is the HTTP status when there is one, and omitted for a timeout —
+ * where there is no response at all, and "the origin did not answer" is the
+ * whole observation. Statuses below 500 are left alone: a 404 from our own
+ * registry for a slug that does not exist is the caller's argument, not our
+ * outage, and marking it would put ordinary 404s on the incident dashboard.
+ *
+ * Idempotent, so a message that is wrapped and re-marked on the way up (the
+ * fleet pack's retry loop re-throws through two layers) carries the marker once.
+ */
+function markInternalOrigin(
+  message: string,
+  url: string | URL | undefined | null,
+  status?: number,
+): string {
+  if (status !== undefined && status < 500) return message;
+  if (!isPipeworxOrigin(url)) return message;
+  if (message.includes(INTERNAL_ORIGIN_MARKER)) return message;
+  return message + INTERNAL_ORIGIN_MARKER;
+}
+
+/**
+ * Which blob4 value a failure from our own web services books as, or undefined
+ * if this is not one.
+ *
+ * Ordered AFTER `internalDbMetricsClass` at the call site: a PostgREST envelope
+ * from our own Supabase is a strictly more specific statement about the same
+ * row (which of our services, and why), and the two cannot disagree about
+ * whether the failure is ours.
+ */
+function internalHostMetricsClass(error: string): string | undefined {
+  return error.includes(INTERNAL_ORIGIN_MARKER) ? INTERNAL_SERVICE_UNREACHABLE_CLASS : undefined;
+}
+/**
+ * MEPS — what Americans actually take, and what for.
+ *
+ * AHRQ's Medical Expenditure Panel Survey is a survey of the whole US civilian
+ * non-institutionalized population: every payer, every age, insured and not.
+ * It is the only source in this catalog that links a prescription to the
+ * CONDITION it was written for.
+ *
+ * These are SURVEY ESTIMATES. Each respondent stands in for ~18,000 Americans,
+ * so a cell backed by three people yields a confident-looking "55,000 people".
+ * Every row therefore carries `sample_persons`, and anything under AHRQ's own
+ * floor of 60 is marked unreliable rather than dressed up or silently dropped.
+ */
+import { resolveConditionCodes } from './conditions.js';
+
+const UA = 'pipeworx/1.0 (+https://pipeworx.io)';
+
+/** AHRQ's reliability floor for a MEPS estimate. */
+const RELIABLE_PERSONS = 60;
+/** Default sample floor for a returned row: AHRQ's own publication bar, so the
+ *  default answer contains nothing AHRQ would decline to publish (Bruce,
+ *  2026-09-09). It was briefly 30, which returned more of the tail — every row
+ *  reported its sample either way, but a caller who does not read
+ *  `sample_persons` was being handed estimates AHRQ would suppress. A caller
+ *  who genuinely wants the tail asks for it with `min_sample` and gets rows
+ *  marked reliable:false. */
+const DEFAULT_MIN_PERSONS = RELIABLE_PERSONS;
+
+const tools: McpToolExport['tools'] = [
+  {
+    name: 'meps_drugs_for_condition',
+    description:
+      'Which prescription drugs Americans actually take for a given medical condition, with a national estimate of how many people take each one. Covers the whole US population — every payer, every age, commercial and uninsured included, not just Medicare or Medicaid. Ask with a plain condition name like "depression", "high cholesterol", "type 2 diabetes" or "asthma". Returns drugs ranked by estimated people treated, each with the survey sample behind it.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        condition: { type: 'string', description: 'Condition in plain words ("depression", "high blood pressure"), or a CCSR category code like MBD002.' },
+        year: { type: 'number', description: 'Survey year. Defaults to the most recent loaded.' },
+        min_sample: { type: 'number', description: `Minimum survey respondents behind a row, default ${DEFAULT_MIN_PERSONS} — AHRQ's own reliability floor. Lower it to see less-common drugs, which come back marked reliable:false because AHRQ would not publish them.` },
+        limit: { type: 'number', description: 'Max drugs, default 20, max 100.' },
+      },
+      required: ['condition'],
+    },
+  },
+  {
+    name: 'meps_conditions_for_drug',
+    description:
+      'What a drug is actually being prescribed for in the US population, ranked by estimated people. Answers off-label and multi-use questions that a label or an approval database cannot: what gabapentin, metformin or amitriptyline are really used to treat in practice. Give a generic drug name.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        drug: { type: 'string', description: 'Generic drug name, e.g. "gabapentin", "metformin".' },
+        year: { type: 'number', description: 'Survey year. Defaults to the most recent loaded.' },
+        min_sample: { type: 'number', description: `Minimum survey respondents behind a row, default ${DEFAULT_MIN_PERSONS}.` },
+        limit: { type: 'number', description: 'Max conditions, default 20, max 100.' },
+      },
+      required: ['drug'],
+    },
+  },
+  {
+    name: 'meps_drug_use',
+    description:
+      'National prescription volume and spending for one drug: how many Americans filled it, how many fills, total spend across all payers and how much patients paid out of pocket. Population-wide rather than a single programme, so it answers how widely used a medicine is in the US and what share of its cost falls on patients.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        drug: { type: 'string', description: 'Generic drug name, e.g. "atorvastatin".' },
+        year: { type: 'number', description: 'Survey year. Defaults to the most recent loaded.' },
+      },
+      required: ['drug'],
+    },
+  },
+  {
+    name: 'meps_top_drugs',
+    description:
+      'The most-used prescription drugs in the United States, ranked by estimated number of people who filled them, with total and out-of-pocket spending. Answers what the most commonly taken medications in America are for a given year.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        year: { type: 'number', description: 'Survey year. Defaults to the most recent loaded.' },
+        order_by: { type: 'string', enum: ['people', 'fills', 'spending'], description: 'Ranking measure, default people.' },
+        limit: { type: 'number', description: 'Max drugs, default 25, max 100.' },
+      },
+    },
+  },
+];
+
+// ---- helpers ---------------------------------------------------------------
+
+const clamp = (v: unknown, def: number, max: number) => {
+  const n = typeof v === 'number' ? v : Number(v);
+  return Number.isFinite(n) ? Math.min(max, Math.max(1, Math.floor(n))) : def;
+};
+
+/** The gateway injects Supabase credentials into `args` (injectSupabase). */
+async function pg(args: Record<string, unknown>, path: string): Promise<any[]> {
+  const url = (args._supabaseUrl as string | undefined)?.trim();
+  const key = (args._supabaseKey as string | undefined)?.trim();
+  if (!url || !key) throw new Error('backing-store credentials not injected');
+  // Bounded: an unbounded fetch to a degraded backing store does not error, it
+  // holds the Worker until its own budget kills the request minutes later.
+  const res = await fetchWithTimeout(
+    `${url}/rest/v1/${path}`,
+    { headers: { apikey: key, Authorization: `Bearer ${key}`, Accept: 'application/json', 'User-Agent': UA } },
+    'MEPS backing store',
+  );
+  // httpError summarises the body rather than forwarding it: a backing-store
+  // error page pasted into a caller-facing message is how a whole HTML document
+  // ends up inside a JSON field.
+  if (!res.ok) throw await httpError(res, 'MEPS backing store');
+  return res.json();
+}
+
+/** Newest year with a COMPLETED load. Reading max(year) from the data instead
+ *  would happily return a year whose ingest died halfway. */
+async function latestYear(args: Record<string, unknown>): Promise<number | null> {
+  const rows = await pg(args, 'meps_ingest_runs?status=eq.ok&select=year&order=year.desc&limit=1');
+  return rows[0]?.year ?? null;
+}
+
+async function resolveYear(args: Record<string, unknown>): Promise<{ year: number | null; requested?: number }> {
+  const asked = typeof args.year === 'number' ? args.year : undefined;
+  if (asked) {
+    const rows = await pg(args, `meps_ingest_runs?status=eq.ok&year=eq.${asked}&select=year&limit=1`);
+    if (rows.length) return { year: asked, requested: asked };
+    return { year: await latestYear(args), requested: asked };
+  }
+  return { year: await latestYear(args) };
+}
+
+/** MEPS reports a therapeutic CLASS in place of a product name when naming the
+ *  product would identify a respondent — "INTERLEUKIN INHIBITORS" rather than a
+ *  specific biologic. 49 of 554 names in 2024 are classes and they carry 31.9%
+ *  of all spending, so a spending ranking is mostly classes at the top. Said
+ *  plainly in the response, because a caller reading "INTERLEUKIN INHIBITORS is
+ *  the top drug by spending" has been misled by our vocabulary, not theirs. */
+const CLASS_NAME_NOTE =
+  'Some names are therapeutic CLASSES, not products — MEPS substitutes the class ' +
+  '("INTERLEUKIN INHIBITORS", "ANTINEOPLASTICS") where naming the drug would identify a ' +
+  'respondent. In 2024 that is 49 of 554 names but 32% of all spending, so class rows ' +
+  'dominate any ranking by cost. Treat a plural or category-sounding name as a class.';
+
+const num = (v: unknown) => (v === null || v === undefined ? null : Math.round(Number(v)));
+const esc = (v: string) => encodeURIComponent(`*${v.replace(/[*,()]/g, ' ').trim()}*`);
+
+/** The shared caveat. Repeated on every response on purpose: a weighted survey
+ *  number read without it is indistinguishable from a census count. */
+function basis(year: number, coverage?: number | null) {
+  return {
+    source: 'AHRQ Medical Expenditure Panel Survey (MEPS), household component',
+    year,
+    population: 'US civilian non-institutionalized population, all payers',
+    estimate_basis:
+      `Weighted survey estimates, not counts. sample_persons is the number of real respondents behind each row; rows below ${RELIABLE_PERSONS} respondents are marked reliable:false, which is AHRQ's own floor.`,
+    ...(coverage != null
+      ? {
+          condition_link_coverage:
+            `${(coverage * 100).toFixed(1)}% of prescription fills in this year are linked to a condition; the rest are not attributed to any diagnosis, so condition totals are a subset of all prescribing.`,
+        }
+      : {}),
+  };
+}
+
+async function coverageFor(args: Record<string, unknown>, year: number): Promise<number | null> {
+  const rows = await pg(args, `meps_ingest_runs?status=eq.ok&year=eq.${year}&select=condition_link_coverage&order=finished_at.desc&limit=1`);
+  return rows[0]?.condition_link_coverage ?? null;
+}
+
+async function labelsFor(args: Record<string, unknown>, codes: string[]): Promise<Record<string, string>> {
+  if (!codes.length) return {};
+  const rows = await pg(args, `meps_ccsr_category?ccsr=in.(${codes.join(',')})&select=ccsr,label`);
+  return Object.fromEntries(rows.map((r: any) => [r.ccsr, r.label]));
+}
+
+// ---- tools -----------------------------------------------------------------
+
+async function drugsForCondition(args: Record<string, unknown>) {
+  const raw = typeof args.condition === 'string' ? args.condition.trim() : '';
+  if (!raw) return { found: false, reason: 'no_condition', hint: 'Give a condition, e.g. "depression" or a CCSR code like MBD002.' };
+
+  const { year, requested } = await resolveYear(args);
+  if (!year) return { found: false, reason: 'no_data_loaded', hint: 'No MEPS year has finished loading yet.' };
+
+  let codes = resolveConditionCodes(raw);
+  let matchedBy: 'code' | 'synonym' | 'label' = codes ? (/^[A-Z]{3}\d{3}$/.test(raw.toUpperCase()) ? 'code' : 'synonym') : 'label';
+  if (!codes) {
+    const hits = await pg(args, `meps_ccsr_category?label=ilike.${esc(raw)}&select=ccsr,label&limit=6`);
+    if (!hits.length) {
+      return {
+        found: false,
+        reason: 'unknown_condition',
+        hint: `No MEPS condition category matches "${raw}". Try a broader word ("diabetes", "asthma", "depression"), or a CCSR code like END010.`,
+      };
+    }
+    codes = hits.map((h: any) => h.ccsr);
+  }
+
+  const minSample = clamp(args.min_sample, DEFAULT_MIN_PERSONS, 100000);
+  const limit = clamp(args.limit, 20, 100);
+  const rows = await pg(
+    args,
+    `meps_rx_condition_drug?year=eq.${year}&ccsr=in.(${codes.join(',')})&persons_unweighted=gte.${minSample}` +
+      `&select=ccsr,drug,persons_unweighted,fills_unweighted,persons_weighted,fills_weighted,reliable` +
+      `&order=persons_weighted.desc&limit=${limit}`
+  );
+
+  const [labels, coverage] = await Promise.all([labelsFor(args, codes), coverageFor(args, year)]);
+
+  if (!rows.length) {
+    // Distinguish "nothing meets the bar" from "nothing at all" — they call for
+    // different next moves, and collapsing them reads as absence of the drug.
+    const anyRows = await pg(args, `meps_rx_condition_drug?year=eq.${year}&ccsr=in.(${codes.join(',')})&select=drug&limit=1`);
+    return {
+      found: false,
+      reason: anyRows.length ? 'below_min_sample' : 'no_match',
+      hint: anyRows.length
+        ? `Rows exist for ${codes.join(', ')} in ${year} but none has ${minSample}+ respondents behind it. Lower min_sample to see them, treating the estimates as weak.`
+        : `MEPS ${year} records no prescriptions linked to ${codes.join(', ')}. Fewer than half of fills carry a condition link at all, so this can mean "not captured" rather than "not prescribed".`,
+      matched_conditions: codes.map((c) => ({ ccsr: c, label: labels[c] ?? c })),
+      ...basis(year, coverage),
+    };
+  }
+
+  return {
+    found: true,
+    count: rows.length,
+    matched_by: matchedBy,
+    matched_conditions: codes.map((c) => ({ ccsr: c, label: labels[c] ?? c })),
+    ...(requested && requested !== year ? { note_year: `MEPS ${requested} is not loaded; answering for ${year}.` } : {}),
+    drugs: rows.map((r: any) => ({
+      drug: r.drug,
+      condition: labels[r.ccsr] ?? r.ccsr,
+      ccsr: r.ccsr,
+      people_estimate: num(r.persons_weighted),
+      fills_estimate: num(r.fills_weighted),
+      sample_persons: r.persons_unweighted,
+      reliable: r.reliable,
+    })),
+    ...basis(year, coverage),
+  };
+}
+
+async function conditionsForDrug(args: Record<string, unknown>) {
+  const drug = typeof args.drug === 'string' ? args.drug.trim().toUpperCase() : '';
+  if (!drug) return { found: false, reason: 'no_drug', hint: 'Give a generic drug name, e.g. "gabapentin".' };
+
+  const { year, requested } = await resolveYear(args);
+  if (!year) return { found: false, reason: 'no_data_loaded', hint: 'No MEPS year has finished loading yet.' };
+
+  const minSample = clamp(args.min_sample, DEFAULT_MIN_PERSONS, 100000);
+  const limit = clamp(args.limit, 20, 100);
+  const rows = await pg(
+    args,
+    `meps_rx_condition_drug?year=eq.${year}&drug=ilike.${esc(drug)}&persons_unweighted=gte.${minSample}` +
+      `&select=ccsr,drug,persons_unweighted,persons_weighted,fills_weighted,reliable&order=persons_weighted.desc&limit=${limit}`
+  );
+  const coverage = await coverageFor(args, year);
+
+  if (!rows.length) {
+    const known = await pg(args, `meps_rx_drug?year=eq.${year}&drug=ilike.${esc(drug)}&select=drug,persons_unweighted&limit=3`);
+    return {
+      found: false,
+      reason: known.length ? 'no_condition_link' : 'unknown_drug',
+      hint: known.length
+        ? `"${known[0].drug}" is in MEPS ${year} but no condition it was prescribed for clears ${minSample} respondents. Fewer than half of fills carry a condition link at all.`
+        : `No drug matching "${drug}" in MEPS ${year}. MEPS records cleaned GENERIC names, so try "atorvastatin" rather than "Lipitor".`,
+      ...basis(year, coverage),
+    };
+  }
+
+  const labels = await labelsFor(args, [...new Set(rows.map((r: any) => r.ccsr))]);
+  return {
+    found: true,
+    drug: rows[0].drug,
+    count: rows.length,
+    ...(requested && requested !== year ? { note_year: `MEPS ${requested} is not loaded; answering for ${year}.` } : {}),
+    conditions: rows.map((r: any) => ({
+      condition: labels[r.ccsr] ?? r.ccsr,
+      ccsr: r.ccsr,
+      people_estimate: num(r.persons_weighted),
+      fills_estimate: num(r.fills_weighted),
+      sample_persons: r.persons_unweighted,
+      reliable: r.reliable,
+    })),
+    ...basis(year, coverage),
+    note: 'A single fill can be linked to more than one condition, so these do not sum to the drug total in meps_drug_use.',
+  };
+}
+
+async function drugUse(args: Record<string, unknown>) {
+  const drug = typeof args.drug === 'string' ? args.drug.trim().toUpperCase() : '';
+  if (!drug) return { found: false, reason: 'no_drug', hint: 'Give a generic drug name, e.g. "atorvastatin".' };
+
+  const { year, requested } = await resolveYear(args);
+  if (!year) return { found: false, reason: 'no_data_loaded', hint: 'No MEPS year has finished loading yet.' };
+
+  const rows = await pg(
+    args,
+    `meps_rx_drug?year=eq.${year}&drug=ilike.${esc(drug)}&select=*&order=persons_weighted.desc&limit=5`
+  );
+  if (!rows.length) {
+    return {
+      found: false,
+      reason: 'unknown_drug',
+      hint: `No drug matching "${drug}" in MEPS ${year}. MEPS records cleaned GENERIC names, so try "atorvastatin" rather than "Lipitor".`,
+      ...basis(year),
+    };
+  }
+
+  const r = rows[0];
+  const total = Number(r.total_expenditure);
+  const oop = Number(r.oop_expenditure);
+  return {
+    found: true,
+    drug: r.drug,
+    ...(requested && requested !== year ? { note_year: `MEPS ${requested} is not loaded; answering for ${year}.` } : {}),
+    people_estimate: num(r.persons_weighted),
+    fills_estimate: num(r.fills_weighted),
+    sample_persons: r.persons_unweighted,
+    reliable: r.reliable,
+    total_spending_usd: num(total),
+    out_of_pocket_usd: num(oop),
+    out_of_pocket_share: Number.isFinite(total) && total > 0 ? Number((oop / total).toFixed(3)) : null,
+    ...(rows.length > 1 ? { other_matches: rows.slice(1).map((x: any) => x.drug) } : {}),
+    ...basis(year),
+    note: CLASS_NAME_NOTE,
+  };
+}
+
+async function topDrugs(args: Record<string, unknown>) {
+  const { year, requested } = await resolveYear(args);
+  if (!year) return { found: false, reason: 'no_data_loaded', hint: 'No MEPS year has finished loading yet.' };
+
+  const orderArg = typeof args.order_by === 'string' ? args.order_by.toLowerCase() : 'people';
+  const ORDER: Record<string, string> = {
+    people: 'persons_weighted',
+    fills: 'fills_weighted',
+    spending: 'total_expenditure',
+  };
+  const col = ORDER[orderArg];
+  if (!col) {
+    return { found: false, reason: 'unknown_order_by', hint: `order_by must be one of: ${Object.keys(ORDER).join(', ')}.` };
+  }
+
+  const limit = clamp(args.limit, 25, 100);
+  const rows = await pg(
+    args,
+    `meps_rx_drug?year=eq.${year}&select=drug,persons_unweighted,persons_weighted,fills_weighted,total_expenditure,oop_expenditure,reliable&order=${col}.desc&limit=${limit}`
+  );
+  if (!rows.length) return { found: false, reason: 'no_match', hint: `No drug rows for MEPS ${year}.`, ...basis(year) };
+
+  return {
+    found: true,
+    count: rows.length,
+    ranked_by: orderArg,
+    ...(requested && requested !== year ? { note_year: `MEPS ${requested} is not loaded; answering for ${year}.` } : {}),
+    drugs: rows.map((r: any) => ({
+      drug: r.drug,
+      people_estimate: num(r.persons_weighted),
+      fills_estimate: num(r.fills_weighted),
+      total_spending_usd: num(r.total_expenditure),
+      out_of_pocket_usd: num(r.oop_expenditure),
+      sample_persons: r.persons_unweighted,
+      reliable: r.reliable,
+    })),
+    ...basis(year),
+    note: CLASS_NAME_NOTE,
+  };
+}
+
+const callTool: McpToolExport['callTool'] = async (name, args) => {
+  switch (name) {
+    case 'meps_drugs_for_condition': return drugsForCondition(args);
+    case 'meps_conditions_for_drug': return conditionsForDrug(args);
+    case 'meps_drug_use': return drugUse(args);
+    case 'meps_top_drugs': return topDrugs(args);
+    default: return { error: `unknown tool: ${name}` };
+  }
+};
+
+export default { tools, callTool } satisfies McpToolExport;