euparliamentmonitor 0.9.10 → 0.9.12

package/README.md

@@ -136,7 +136,7 @@ The published site is the audience-facing companion to this npm/TypeScript packa
 
 **MCP Server Integration**: The project uses the
 [European-Parliament-MCP-Server](https://github.com/Hack23/European-Parliament-MCP-Server)
-v1.3.3 for accessing real EU Parliament data via the Model Context Protocol.
+v1.3.6 for accessing real EU Parliament data via the Model Context Protocol.
 
 - **MCP Server Status**: ✅ Fully operational — 60+ EP data tools available
   (feeds, direct lookups, analytical tools, intelligence correlation)
@@ -432,7 +432,7 @@ import type { ArticleCategory, LanguageCode } from 'euparliamentmonitor/types';
 
 ## 🔌 Data Sources
 
-**Primary — European Parliament MCP Server** ([Hack23/European-Parliament-MCP-Server](https://github.com/Hack23/European-Parliament-MCP-Server) v1.3.3+, fully operational):
+**Primary — European Parliament MCP Server** ([Hack23/European-Parliament-MCP-Server](https://github.com/Hack23/European-Parliament-MCP-Server) v1.3.6+, fully operational):
 
 - 🗳️ Plenary sessions, voting records, roll-call votes
 - 📜 Adopted texts, motions, resolutions, urgency files

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "euparliamentmonitor",
-  "version": "0.9.10",
+  "version": "0.9.12",
   "type": "module",
   "description": "European Parliament Intelligence Platform - Monitor political activity with systematic transparency",
   "main": "scripts/index.js",
@@ -74,6 +74,7 @@
     "optimize-css": "node scripts/optimize-css.js",
     "minify-assets": "node scripts/minify-assets.js",
     "validate-ep-api": "npx tsx src/utils/validate-ep-api.ts",
+    "mcp:probe": "npx tsx src/utils/mcp-probe.ts",
     "lint:prompts": "node scripts/lint-prompts.js",
     "htmlhint": "sh -c 'htmlhint *.html; set -- news/*.html; if [ -e \"$1\" ]; then htmlhint \"$@\"; else echo \"No news/*.html files to lint\"; fi'",
     "serve": "python3 -m http.server 8080",
@@ -149,7 +150,7 @@
     "@playwright/test": "1.60.0",
     "@types/d3": "7.4.3",
     "@types/markdown-it": "^14.1.2",
-    "@types/node": "25.7.0",
+    "@types/node": "25.8.0",
     "@types/papaparse": "5.5.2",
     "@typescript-eslint/eslint-plugin": "8.59.3",
     "@typescript-eslint/parser": "8.59.3",
@@ -159,7 +160,7 @@
     "chartjs-plugin-annotation": "3.1.0",
     "clean-css": "^5.3.3",
     "d3": "7.9.0",
-    "eslint": "10.3.0",
+    "eslint": "10.4.0",
     "eslint-config-prettier": "10.1.8",
     "eslint-plugin-jsdoc": "62.9.0",
     "eslint-plugin-security": "4.0.0",
@@ -168,7 +169,7 @@
     "html-minifier-terser": "^7.2.0",
     "htmlhint": "1.9.2",
     "husky": "9.1.7",
-    "jscpd": "4.1.1",
+    "jscpd": "4.2.1",
     "knip": "^6.7.0",
     "lint-staged": "17.0.4",
     "mermaid": "11.15.0",
@@ -178,7 +179,7 @@
     "sharp": "^0.34.5",
     "terser": "^5.47.1",
     "ts-api-utils": "2.5.0",
-    "tsx": "4.21.0",
+    "tsx": "4.22.0",
     "typedoc": "0.28.19",
     "typescript": "6.0.3",
     "vitest": "4.1.6"
@@ -187,7 +188,7 @@
     "node": ">=26"
   },
   "dependencies": {
-    "european-parliament-mcp-server": "1.3.3",
+    "european-parliament-mcp-server": "1.3.6",
     "markdown-it": "^14.1.1",
     "markdown-it-anchor": "^9.2.0",
     "markdown-it-attrs": "^4.3.1",

package/scripts/cache-thresholds.js

@@ -0,0 +1,305 @@
+#!/usr/bin/env node
+// SPDX-FileCopyrightText: 2024-2026 Hack23 AB
+// SPDX-License-Identifier: Apache-2.0
+
+/**
+ * @module scripts/cache-thresholds
+ * @description Thresholds cache helper for Stage B analysis runs.
+ *
+ * The `analysis/methodologies/reference-quality-thresholds.json` file is large
+ * (~1 000 lines). Without this helper, each artifact's Stage B write cycle
+ * causes the agent to re-read the entire thresholds file to look up the floor
+ * for a single relativePath. With 38+ artifacts per run, that amounts to 38+
+ * wasted LLM invocations.
+ *
+ * This script:
+ *   1. Reads the full `reference-quality-thresholds.json`
+ *   2. Filters to the subset relevant to the given article-type slug
+ *   3. Writes a compact `thresholds-cache.json` under the run directory
+ *   4. Uses a content hash for the cache key so subsequent calls with identical
+ *      inputs short-circuit without re-writing the file
+ *
+ * The output file is a flat map: `{ [relativePath]: number }` where the value
+ * is the per-artifact line floor for that slug, plus a small set of metadata
+ * fields.
+ *
+ * Invocation (Stage B start — called once per run):
+ *   node scripts/cache-thresholds.js \
+ *     --slug breaking \
+ *     --run-id breaking-run-1234 \
+ *     [--analysis-dir analysis/daily/2026-05-14/breaking] \
+ *     [--repo-root /path/to/repo]
+ *
+ * Exports (for unit testing):
+ *   loadThresholds(repoRoot)
+ *   filterThresholdsForSlug(thresholds, slug)
+ *   buildThresholdsCache(repoRoot, slug, runId, runDir)
+ */
+
+import crypto from 'node:crypto';
+import fs from 'node:fs';
+import path from 'node:path';
+import process from 'node:process';
+import { fileURLToPath } from 'node:url';
+
+/** Path relative to repo root for the canonical thresholds file. */
+const THRESHOLDS_REL_PATH = 'analysis/methodologies/reference-quality-thresholds.json';
+
+/** Default minimum line floor when no per-artifact entry is found. */
+const DEFAULT_MIN_LINES = 30;
+
+/** Output filename written to the run directory. */
+const CACHE_FILENAME = 'thresholds-cache.json';
+
+// ---------------------------------------------------------------------------
+// Article-type slug → thresholds key normalisation
+// ---------------------------------------------------------------------------
+
+/**
+ * Normalise an article-type slug to the canonical key used in the thresholds
+ * JSON. The slug might be a workflow slug (e.g. `breaking`) or a longer form
+ * (e.g. `breaking-run-123456`). We strip any run-ID suffix.
+ *
+ * @param {string} slug - Raw slug value
+ * @returns {string} Canonical article-type key
+ */
+export function normaliseSlug(slug) {
+  // Strip run-ID suffixes in either form:
+  //   - `-run-<digits>` (e.g. `breaking-run-1234`) — newer workflow runs
+  //   - `-run<digits>`  (e.g. `breaking-run184`)   — legacy folder names
+  // Each form may carry an optional `-<digits>` sub-run suffix used by
+  // long-horizon workflows (e.g. `breaking-run-123-456`).
+  return slug.replace(/-run-?\d+(-\d+)?$/, '');
+}
+
+// ---------------------------------------------------------------------------
+// Thresholds file loader
+// ---------------------------------------------------------------------------
+
+/**
+ * Load the full `reference-quality-thresholds.json` from the repo root.
+ *
+ * @param {string} repoRoot - Absolute path to the repository root
+ * @returns {{ thresholds: object, raw: string }}
+ * @throws {Error} If the file is missing or not valid JSON
+ */
+export function loadThresholds(repoRoot) {
+  const filePath = path.join(repoRoot, THRESHOLDS_REL_PATH);
+
+  if (!fs.existsSync(filePath)) {
+    throw new Error(`Thresholds file not found: ${filePath}`);
+  }
+
+  const raw = fs.readFileSync(filePath, 'utf8');
+  const parsed = JSON.parse(raw);
+  return { thresholds: parsed, raw };
+}
+
+// ---------------------------------------------------------------------------
+// Slug filter
+// ---------------------------------------------------------------------------
+
+/**
+ * Filter the thresholds object to only the entries relevant to a given slug.
+ *
+ * The thresholds JSON has a `thresholds` object keyed by `articleType`.
+ * For each article type, it contains a `files` map of `relativePath → number`.
+ *
+ * The output is a flat map: `{ [relativePath]: floor }` for the matching
+ * article type(s). Cross-type entries (present in multiple article types)
+ * use the maximum floor across all matched types.
+ *
+ * @param {object} thresholds - Parsed thresholds JSON root object
+ * @param {string} slug       - Article-type slug (already normalised)
+ * @returns {{ floors: Record<string, number>, matchedTypes: string[], defaultFloor: number }}
+ */
+export function filterThresholdsForSlug(thresholds, slug) {
+  const floors = {};
+  const matchedTypes = [];
+
+  const thresholdsMap = thresholds?.thresholds ?? {};
+  const defaultFloor =
+    thresholds?.defaults?.minLines ?? DEFAULT_MIN_LINES;
+
+  for (const [articleType, typeConfig] of Object.entries(thresholdsMap)) {
+    // Match when the articleType equals the slug or contains it as a prefix
+    const normType = normaliseSlug(articleType);
+    if (normType !== slug && articleType !== slug) continue;
+
+    matchedTypes.push(articleType);
+
+    const filesMap = typeConfig?.files ?? typeConfig ?? {};
+    for (const [relPath, floor] of Object.entries(filesMap)) {
+      if (typeof floor === 'number') {
+        // Use max floor across matched types for cross-type overlap
+        floors[relPath] = Math.max(floors[relPath] ?? 0, floor);
+      }
+    }
+  }
+
+  return { floors, matchedTypes, defaultFloor };
+}
+
+// ---------------------------------------------------------------------------
+// Cache builder
+// ---------------------------------------------------------------------------
+
+/**
+ * Build and write a slug-filtered thresholds cache file.
+ *
+ * If an existing cache file with the same content hash already exists,
+ * the write is skipped and `{ cached: true }` is returned.
+ *
+ * @param {string} repoRoot   - Absolute path to the repository root
+ * @param {string} slug       - Article-type slug
+ * @param {string} runId      - Run identifier (used for the output path)
+ * @param {string} [runDir]   - Override for the run directory (defaults to
+ *                              `analysis/runs/<runId>` relative to repoRoot)
+ * @returns {{
+ *   outputFile: string,
+ *   floors: Record<string, number>,
+ *   matchedTypes: string[],
+ *   defaultFloor: number,
+ *   contentHash: string,
+ *   cached: boolean
+ * }}
+ */
+export function buildThresholdsCache(repoRoot, slug, runId, runDir) {
+  const normSlug = normaliseSlug(slug);
+
+  const { thresholds, raw } = loadThresholds(repoRoot);
+
+  // Content hash of the source thresholds file (not the filtered output).
+  const contentHash = crypto.createHash('sha256').update(raw).digest('hex').slice(0, 16);
+
+  const { floors, matchedTypes, defaultFloor } = filterThresholdsForSlug(thresholds, normSlug);
+
+  const outputDir = runDir
+    ? runDir
+    : path.join(repoRoot, 'analysis', 'runs', runId);
+
+  fs.mkdirSync(outputDir, { recursive: true });
+
+  const outputFile = path.join(outputDir, CACHE_FILENAME);
+
+  const payload = {
+    generatedAt: new Date().toISOString(),
+    slug: normSlug,
+    runId,
+    contentHash,
+    defaultFloor,
+    matchedTypes,
+    artifactCount: Object.keys(floors).length,
+    floors,
+  };
+
+  const payloadJson = JSON.stringify(payload, null, 2);
+
+  // Short-circuit if file already exists with identical content hash
+  if (fs.existsSync(outputFile)) {
+    try {
+      const existing = JSON.parse(fs.readFileSync(outputFile, 'utf8'));
+      if (existing.contentHash === contentHash && existing.slug === normSlug) {
+        return { outputFile, floors, matchedTypes, defaultFloor, contentHash, cached: true };
+      }
+    } catch {
+      // Fall through to (re)write
+    }
+  }
+
+  fs.writeFileSync(outputFile, payloadJson, 'utf8');
+
+  return { outputFile, floors, matchedTypes, defaultFloor, contentHash, cached: false };
+}
+
+// ---------------------------------------------------------------------------
+// CLI entry point
+// ---------------------------------------------------------------------------
+
+/**
+ * Parse minimalist `--key value` CLI args.
+ *
+ * @param {string[]} argv
+ * @returns {Record<string, string|boolean>}
+ */
+/* c8 ignore start */
+function parseArgs(argv) {
+  const out = {};
+  let i = 0;
+  while (i < argv.length) {
+    const arg = argv[i];
+    if (arg.startsWith('--')) {
+      const key = arg.slice(2);
+      const next = argv[i + 1];
+      if (next === undefined || next.startsWith('--')) {
+        out[key] = true;
+        i += 1;
+      } else {
+        out[key] = next;
+        i += 2;
+      }
+    } else {
+      i += 1;
+    }
+  }
+  return out;
+}
+
+/**
+ * CLI main entry point.
+ *
+ * @param {string[]} [argv]
+ * @returns {void}
+ */
+export function main(argv = process.argv.slice(2)) {
+  const args = parseArgs(argv);
+
+  if (!args.slug || !args['run-id']) {
+    process.stderr.write(
+      'Usage: node scripts/cache-thresholds.js --slug <slug> --run-id <run-id>' +
+        ' [--analysis-dir <dir>] [--repo-root <path>]\n',
+    );
+    process.exit(2);
+  }
+
+  const slug = String(args.slug);
+  const runId = String(args['run-id']);
+  const repoRoot = args['repo-root']
+    ? String(args['repo-root'])
+    : process.cwd();
+  const runDir = args['analysis-dir']
+    ? path.join(String(args['analysis-dir']), 'runs', runId)
+    : undefined;
+
+  try {
+    const result = buildThresholdsCache(repoRoot, slug, runId, runDir);
+
+    process.stdout.write(
+      JSON.stringify({
+        status: 'ok',
+        outputFile: result.outputFile,
+        slug,
+        matchedTypes: result.matchedTypes,
+        artifactCount: Object.keys(result.floors).length,
+        defaultFloor: result.defaultFloor,
+        contentHash: result.contentHash,
+        cached: result.cached,
+      }) + '\n',
+    );
+  } catch (err) {
+    process.stderr.write(`Error: ${err}\n`);
+    process.exit(1);
+  }
+}
+
+// Standard ESM CLI guard
+const isMain =
+  typeof process !== 'undefined' &&
+  process.argv[1] !== undefined &&
+  (process.argv[1] === fileURLToPath(import.meta.url) ||
+    process.argv[1].endsWith('/cache-thresholds.js'));
+
+if (isMain) {
+  main();
+}
+/* c8 ignore stop */