@jackwener/opencli 1.7.11 → 1.7.12

package/clis/indeed/search.js

@@ -0,0 +1,110 @@
+/**
+ * Indeed keyword search.
+ *
+ * Drives the public `/jobs?q=&l=` listing page through a real browser
+ * because Cloudflare returns 403 to bare fetches. Extraction happens
+ * inside `page.evaluate` against the rendered DOM (Indeed's JSON-island
+ * is heavily obfuscated and changes shape per A/B bucket).
+ */
+
+import { cli, Strategy } from '@jackwener/opencli/registry';
+import { EmptyResultError, CommandExecutionError } from '@jackwener/opencli/errors';
+import {
+    SEARCH_COLUMNS,
+    requireQuery,
+    requireBoundedInt,
+    requireNonNegativeInt,
+    requireFromage,
+    requireSort,
+    buildSearchUrl,
+    searchCardToRow,
+} from './utils.js';
+
+cli({
+    site: 'indeed',
+    name: 'search',
+    access: 'read',
+    description: 'Indeed keyword job search (rendered DOM via browser session, US site)',
+    domain: 'www.indeed.com',
+    strategy: Strategy.COOKIE,
+    browser: true,
+    navigateBefore: false,
+    args: [
+        { name: 'query', positional: true, required: true, help: 'Job keyword (title / skill / company)' },
+        { name: 'location', type: 'string', default: '', help: 'Location filter (e.g. "remote", "New York, NY", "San Francisco")' },
+        { name: 'fromage', type: 'string', default: '', help: 'Recency filter, days back: 1 / 3 / 7 / 14' },
+        { name: 'sort', type: 'string', default: 'relevance', help: 'Sort order: relevance | date' },
+        { name: 'start', type: 'int', default: 0, help: 'Pagination offset (multiple of 10, 0-based)' },
+        { name: 'limit', type: 'int', default: 15, help: 'Max rows to return (1-25, capped at one page)' },
+    ],
+    columns: SEARCH_COLUMNS,
+    func: async (page, kwargs) => {
+        const query = requireQuery(kwargs.query);
+        const location = String(kwargs.location ?? '').trim();
+        const fromage = requireFromage(kwargs.fromage);
+        const sort = requireSort(kwargs.sort);
+        const start = requireNonNegativeInt(kwargs.start, 0, 'start');
+        const limit = requireBoundedInt(kwargs.limit, 15, 25, 'limit');
+
+        const url = buildSearchUrl({ query, location, fromage, sort, start });
+        await page.goto(url);
+        await page.wait(4);
+
+        let cards;
+        try {
+            cards = await page.evaluate(`(async () => {
+                const hasResults = () => !!document.querySelector('.job_seen_beacon');
+                const hasEmptyState = () => {
+                    const text = document.body?.innerText || '';
+                    return !!document.querySelector('[data-testid="searchCountPages"], [data-testid="searchCount"], [data-testid="noResultsMessage"], [data-testid="empty-serp-result"]')
+                        || /did not match any jobs|no jobs found|0 jobs/i.test(text);
+                };
+                let ready = hasResults() || hasEmptyState();
+                for (let i = 0; i < 30; i++) {
+                    if (ready) break;
+                    await new Promise(r => setTimeout(r, 500));
+                    ready = hasResults() || hasEmptyState();
+                }
+                const blocks = document.querySelectorAll('.job_seen_beacon');
+                const seen = new Set();
+                const out = [];
+                for (const b of blocks) {
+                    const titleA = b.querySelector('h2.jobTitle a, [class*="jcs-JobTitle"]');
+                    const jk = titleA?.getAttribute('data-jk');
+                    if (!jk || seen.has(jk)) continue;
+                    seen.add(jk);
+                    const tags = Array.from(b.querySelectorAll('.metadataContainer li span'))
+                        .map(s => (s.textContent || '').trim())
+                        .filter(Boolean);
+                    out.push({
+                        jk,
+                        title: b.querySelector('h2.jobTitle span')?.textContent?.trim() ?? '',
+                        company: b.querySelector('[data-testid="company-name"]')?.textContent?.trim() ?? '',
+                        location: b.querySelector('[data-testid="text-location"]')?.textContent?.trim() ?? '',
+                        salary: b.querySelector('.salary-snippet-container span')?.textContent?.trim() ?? '',
+                        tags,
+                    });
+                }
+                const blockedHeadline = document.title || '';
+                const challenge = blockedHeadline.includes('Just a moment') || !!document.querySelector('[id^="cf-"]');
+                return { cards: out, challenge, ready };
+            })()`);
+        }
+        catch (e) {
+            throw new CommandExecutionError(`Failed to scrape Indeed search DOM: ${e?.message ?? e}`, 'The page may not have fully loaded; try again.');
+        }
+
+        if (cards?.challenge) {
+            throw new CommandExecutionError('Indeed served a Cloudflare challenge page', 'Open https://www.indeed.com in the connected browser and clear the challenge, then retry.');
+        }
+        if (!cards?.ready) {
+            throw new CommandExecutionError('Indeed search page did not expose result or empty-state markers within 15s', 'Indeed may still be loading or the DOM shape may have changed; retry after opening Indeed in the connected browser.');
+        }
+
+        const list = Array.isArray(cards?.cards) ? cards.cards : [];
+        if (list.length === 0) {
+            throw new EmptyResultError('indeed search', `No Indeed jobs matched "${query}"${location ? ` in ${location}` : ''}`);
+        }
+        return list.slice(0, limit).map((c, i) => searchCardToRow(c, start + i + 1));
+    },
+});

package/clis/indeed/utils.js

@@ -0,0 +1,152 @@
+/**
+ * Indeed adapter utilities.
+ *
+ * Indeed sits behind Cloudflare and answers bare fetches with HTTP 403
+ * (`cf-mitigated: challenge`). The whole adapter therefore runs through a
+ * real browser session (Strategy.COOKIE), and DOM extraction lives inside
+ * `page.evaluate` blocks. The helpers below are mostly arg-validation +
+ * pure normalizers so they stay unit-testable without a browser.
+ */
+
+import { ArgumentError } from '@jackwener/opencli/errors';
+
+export const INDEED_ORIGIN = 'https://www.indeed.com';
+
+/** Job key (jk) shape — 16-char lowercase hex. */
+const JK_PATTERN = /^[a-f0-9]{16}$/;
+
+const FROMAGE_VALUES = new Set(['1', '3', '7', '14']);
+
+const SORT_VALUES = new Set(['relevance', 'date']);
+
+/**
+ * Coerce a value to a strict integer. Accepts numeric strings, rejects
+ * floats / non-numeric / NaN. Returns NaN on invalid input so callers can
+ * decide on the right typed error.
+ */
+export function coerceInt(value) {
+    if (value === undefined || value === null || value === '') return NaN;
+    const n = typeof value === 'number' ? value : Number(value);
+    return Number.isFinite(n) && Number.isInteger(n) ? n : NaN;
+}
+
+export function requireBoundedInt(value, defaultValue, maxValue, label) {
+    const raw = value ?? defaultValue;
+    const n = coerceInt(raw);
+    if (!Number.isInteger(n) || n <= 0) {
+        throw new ArgumentError(`indeed ${label} must be a positive integer`);
+    }
+    if (n > maxValue) {
+        throw new ArgumentError(`indeed ${label} must be <= ${maxValue}`);
+    }
+    return n;
+}
+
+export function requireNonNegativeInt(value, defaultValue, label) {
+    const raw = value ?? defaultValue;
+    const n = coerceInt(raw);
+    if (!Number.isInteger(n) || n < 0) {
+        throw new ArgumentError(`indeed ${label} must be a non-negative integer`);
+    }
+    return n;
+}
+
+export function requireJobKey(value) {
+    const id = String(value ?? '').trim().toLowerCase();
+    if (!id) {
+        throw new ArgumentError('indeed job id is required');
+    }
+    if (!JK_PATTERN.test(id)) {
+        throw new ArgumentError(`indeed job id "${value}" is not a valid jk (expected 16-char lowercase hex)`);
+    }
+    return id;
+}
+
+export function requireQuery(value, label = 'query') {
+    const q = String(value ?? '').trim();
+    if (!q) {
+        throw new ArgumentError(`indeed ${label} cannot be empty`);
+    }
+    return q;
+}
+
+/** "1" / "3" / "7" / "14" — accepted by Indeed's `fromage` filter. */
+export function requireFromage(value) {
+    if (value === undefined || value === null || value === '') return '';
+    const v = String(value).trim();
+    if (!FROMAGE_VALUES.has(v)) {
+        throw new ArgumentError(`indeed fromage must be one of 1/3/7/14 (days), got "${value}"`);
+    }
+    return v;
+}
+
+export function requireSort(value, defaultValue = 'relevance') {
+    const v = String(value ?? defaultValue).trim().toLowerCase();
+    if (!SORT_VALUES.has(v)) {
+        throw new ArgumentError(`indeed sort must be "relevance" or "date", got "${value}"`);
+    }
+    return v;
+}
+
+/**
+ * Build an Indeed search URL with only the user-supplied filters set.
+ * Indeed treats absent params as defaults; we never pass empty strings
+ * because the page will echo them back into the query and the URL
+ * stays cleaner for round-tripping.
+ */
+export function buildSearchUrl({ query, location, fromage, sort, start }) {
+    const params = new URLSearchParams();
+    params.set('q', query);
+    if (location) params.set('l', location);
+    if (fromage) params.set('fromage', fromage);
+    if (sort && sort !== 'relevance') params.set('sort', sort);
+    if (start && start > 0) params.set('start', String(start));
+    return `${INDEED_ORIGIN}/jobs?${params.toString()}`;
+}
+
+export function buildJobUrl(jk) {
+    return `${INDEED_ORIGIN}/viewjob?jk=${jk}`;
+}
+
+/** Strip a salary-snippet duplicate out of the metadata pill list. */
+export function dedupeTags(tags, salary) {
+    const out = [];
+    for (const t of tags) {
+        const trimmed = String(t || '').trim();
+        if (!trimmed) continue;
+        if (salary && trimmed === salary) continue;
+        if (out.includes(trimmed)) continue;
+        out.push(trimmed);
+    }
+    return out.join(' · ');
+}
+
+/**
+ * Normalize a parsed search-card object into the row shape declared by
+ * `SEARCH_COLUMNS`. Drops nulls into empty strings, defends against
+ * Indeed surface drift by treating missing fields as empty rather than
+ * silently mislabeling them.
+ */
+export function searchCardToRow(card, rank) {
+    const jk = String(card?.jk ?? '').trim();
+    const salary = String(card?.salary ?? '').trim();
+    const tags = Array.isArray(card?.tags) ? card.tags : [];
+    return {
+        rank,
+        id: jk,
+        title: String(card?.title ?? '').replace(/\s+/g, ' ').trim(),
+        company: String(card?.company ?? '').replace(/\s+/g, ' ').trim(),
+        location: String(card?.location ?? '').replace(/\s+/g, ' ').trim(),
+        salary,
+        tags: dedupeTags(tags, salary),
+        url: jk ? buildJobUrl(jk) : '',
+    };
+}
+
+export const SEARCH_COLUMNS = [
+    'rank', 'id', 'title', 'company', 'location', 'salary', 'tags', 'url',
+];
+
+export const JOB_COLUMNS = [
+    'id', 'title', 'company', 'location', 'salary', 'job_type', 'description', 'url',
+];

package/clis/instagram/collection-create.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'collection-create',
+    access: 'write',
     description: 'Create a new Instagram saved-posts collection (folder)',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/collection-delete.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'collection-delete',
+    access: 'write',
     description: 'Delete an Instagram saved-posts collection (folder) by name or id',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/comment.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'comment',
+    access: 'write',
     description: 'Comment on an Instagram post',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/download.js

@@ -196,6 +196,7 @@ async function downloadInstagramMedia(items, outputDir) {
 cli({
     site: 'instagram',
     name: 'download',
+    access: 'read',
     description: 'Download images and videos from Instagram posts and reels',
     domain: 'www.instagram.com',
     strategy: Strategy.COOKIE,

package/clis/instagram/explore.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'explore',
+    access: 'read',
     description: 'Instagram explore/discover trending posts',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/follow.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'follow',
+    access: 'write',
     description: 'Follow an Instagram user',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/followers.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'followers',
+    access: 'read',
     description: 'List followers of an Instagram user',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/following.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'following',
+    access: 'read',
     description: 'List accounts an Instagram user is following',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/like.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'like',
+    access: 'write',
     description: 'Like an Instagram post',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/note.js

@@ -198,6 +198,7 @@ function buildPublishInstagramNoteJs(content) {
 cli({
     site: 'instagram',
     name: 'note',
+    access: 'write',
     description: 'Publish a text Instagram note',
     domain: 'www.instagram.com',
     strategy: Strategy.UI,

package/clis/instagram/post.js

@@ -1397,6 +1397,7 @@ async function resolveLatestPostUrl(page, existingPostPaths) {
 cli({
     site: 'instagram',
     name: 'post',
+    access: 'write',
     description: 'Post an Instagram feed image or mixed-media carousel',
     domain: 'www.instagram.com',
     strategy: Strategy.UI,

package/clis/instagram/profile.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'profile',
+    access: 'read',
     description: 'Get Instagram user profile info',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/reel.js

@@ -732,6 +732,7 @@ async function resolveLatestReelUrl(page, existingPaths) {
 cli({
     site: 'instagram',
     name: 'reel',
+    access: 'write',
     description: 'Post an Instagram reel video',
     domain: 'www.instagram.com',
     strategy: Strategy.UI,

package/clis/instagram/save.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'save',
+    access: 'write',
     description: 'Save (bookmark) an Instagram post',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/saved.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'saved',
+    access: 'read',
     description: 'Get your saved Instagram posts (optionally from a specific collection)',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/search.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'search',
+    access: 'read',
     description: 'Search Instagram users',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/story.js

@@ -82,6 +82,7 @@ function buildStorySuccessResult(mediaItem, url) {
 cli({
     site: 'instagram',
     name: 'story',
+    access: 'write',
     description: 'Post a single Instagram story image or video',
     domain: 'www.instagram.com',
     strategy: Strategy.UI,

package/clis/instagram/unfollow.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'unfollow',
+    access: 'write',
     description: 'Unfollow an Instagram user',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/unlike.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'unlike',
+    access: 'write',
     description: 'Unlike an Instagram post',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/unsave.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'unsave',
+    access: 'write',
     description: 'Unsave (remove bookmark) an Instagram post',
     domain: 'www.instagram.com',
     args: [

package/clis/instagram/user.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'instagram',
     name: 'user',
+    access: 'read',
     description: 'Get recent posts from an Instagram user',
     domain: 'www.instagram.com',
     args: [

package/clis/jd/add-cart.js

@@ -4,6 +4,7 @@ import { clampInt, normalizeNumericId } from '../_shared/common.js';
 cli({
     site: 'jd',
     name: 'add-cart',
+    access: 'write',
     description: '京东加入购物车',
     domain: 'item.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jd/cart.js

@@ -3,6 +3,7 @@ import { cli, Strategy } from '@jackwener/opencli/registry';
 cli({
     site: 'jd',
     name: 'cart',
+    access: 'read',
     description: '查看京东购物车',
     domain: 'cart.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jd/detail.js

@@ -3,6 +3,7 @@ import { normalizeNumericId } from '../_shared/common.js';
 cli({
     site: 'jd',
     name: 'detail',
+    access: 'read',
     description: '京东商品详情',
     domain: 'item.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jd/item.js

@@ -543,6 +543,7 @@ function detectJdPageState(expectedSku) {
 cli({
     site: 'jd',
     name: 'item',
+    access: 'read',
     description: '京东商品详情（价格、店铺、规格参数、主图、详情图）',
     domain: 'item.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jd/reviews.js

@@ -3,6 +3,7 @@ import { clampInt, normalizeNumericId } from '../_shared/common.js';
 cli({
     site: 'jd',
     name: 'reviews',
+    access: 'read',
     description: '京东商品评价',
     domain: 'item.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jd/search.js

@@ -3,6 +3,7 @@ import { clampInt, requireNonEmptyQuery } from '../_shared/common.js';
 cli({
     site: 'jd',
     name: 'search',
+    access: 'read',
     description: '京东商品搜索',
     domain: 'search.jd.com',
     strategy: Strategy.COOKIE,

package/clis/jianyu/detail.js

@@ -3,6 +3,7 @@ import { runProcurementDetail } from './shared/procurement-detail.js';
 cli({
     site: 'jianyu',
     name: 'detail',
+    access: 'read',
     description: '读取剑鱼标讯详情页并抽取证据字段',
     domain: 'www.jianyu360.cn',
     strategy: Strategy.COOKIE,

package/clis/jianyu/search.js

@@ -526,6 +526,7 @@ function collectApiRowsFromResponses(responses) {
 cli({
     site: SITE,
     name: 'search',
+    access: 'read',
     description: '搜索剑鱼标讯公告',
     domain: DOMAIN,
     strategy: Strategy.COOKIE,

package/clis/jike/comment.js

@@ -8,6 +8,7 @@ import { cli, Strategy } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'comment',
+    access: 'write',
     description: '评论即刻帖子',
     domain: 'web.okjike.com',
     strategy: Strategy.UI,

package/clis/jike/create.js

@@ -8,6 +8,7 @@ import { cli, Strategy } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'create',
+    access: 'write',
     description: '发布即刻动态',
     domain: 'web.okjike.com',
     strategy: Strategy.UI,

package/clis/jike/feed.js

@@ -9,6 +9,7 @@ import { getPostDataJs } from './utils.js';
 cli({
     site: 'jike',
     name: 'feed',
+    access: 'read',
     description: '即刻首页动态流',
     domain: 'web.okjike.com',
     strategy: Strategy.COOKIE,
@@ -16,7 +17,7 @@ cli({
     args: [
         { name: 'limit', type: 'int', default: 20 },
     ],
-    columns: ['author', 'content', 'likes', 'comments', 'time', 'url'],
+    columns: ['id', 'author', 'content', 'likes', 'comments', 'time', 'url'],
     func: async (page, kwargs) => {
         const limit = kwargs.limit || 20;
         // 1. 导航到即刻首页，等待 SPA 重定向到 /following
@@ -43,6 +44,7 @@ cli({
           if (!author && !content) continue;
 
           results.push({
+            id: data.id,
             author,
             content: content.replace(/\\n/g, ' ').slice(0, 120),
             likes: data.likeCount || 0,

package/clis/jike/like.js

@@ -8,6 +8,7 @@ import { cli, Strategy } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'like',
+    access: 'write',
     description: '点赞即刻帖子',
     domain: 'web.okjike.com',
     strategy: Strategy.UI,

package/clis/jike/notifications.js

@@ -21,6 +21,7 @@ function resolveActionLabel(type) {
 cli({
     site: 'jike',
     name: 'notifications',
+    access: 'read',
     description: '即刻通知',
     domain: 'web.okjike.com',
     strategy: Strategy.COOKIE,

package/clis/jike/post.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'post',
+    access: 'read',
     description: '即刻帖子详情及评论',
     domain: 'm.okjike.com',
     browser: true,

package/clis/jike/repost.js

@@ -9,6 +9,7 @@ import { cli, Strategy } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'repost',
+    access: 'write',
     description: '转发即刻帖子',
     domain: 'web.okjike.com',
     strategy: Strategy.UI,

package/clis/jike/search.js

@@ -9,6 +9,7 @@ import { getPostDataJs } from './utils.js';
 cli({
     site: 'jike',
     name: 'search',
+    access: 'read',
     description: '搜索即刻帖子',
     domain: 'web.okjike.com',
     strategy: Strategy.COOKIE,
@@ -17,7 +18,7 @@ cli({
         { name: 'query', type: 'string', required: true, positional: true },
         { name: 'limit', type: 'int', default: 20 },
     ],
-    columns: ['author', 'content', 'likes', 'comments', 'time', 'url'],
+    columns: ['id', 'author', 'content', 'likes', 'comments', 'time', 'url'],
     func: async (page, kwargs) => {
         const keyword = kwargs.query;
         const limit = kwargs.limit || 20;
@@ -43,6 +44,7 @@ cli({
           if (!author && !content) continue;
 
           results.push({
+            id: data.id,
             author,
             content: content.replace(/\\n/g, ' ').slice(0, 120),
             likes: data.likeCount || 0,

package/clis/jike/topic.js

@@ -2,6 +2,7 @@ import { cli } from '@jackwener/opencli/registry';
 cli({
     site: 'jike',
     name: 'topic',
+    access: 'read',
     description: '即刻话题/圈子帖子',
     domain: 'm.okjike.com',
     browser: true,