Commit Graph

448 Commits

Author SHA1 Message Date
jackwener c77c8a8e3a feat: add Coupang search and add-to-cart adapters
Add browser-backed Coupang adapters for search and add-to-cart workflows.
- coupang search: multi-strategy data collection (API/JSON-LD/bootstrap/DOM), structured fields (price, rating, rocket, delivery), pagination and rocket filter support
- coupang add-to-cart: logged-in browser session reuse, stops before checkout
- coupang.ts: comprehensive data normalization layer with badge/rocket/delivery mapping
- browser-tab.ts: withTemporaryTab utility for isolated tab operations
- Unit tests for core normalization functions

Co-authored-by: CodeBBakGoSu <127713112+CodeBBakGoSu@users.noreply.github.com>
2026-03-16 13:30:42 +08:00
jackwener 9e024d4e46 0.5.2 v0.5.2 2026-03-16 13:22:18 +08:00
jackwener 34a9bff2b3 refactor: eliminate code duplication, improve type safety, add tests
- NEW: src/interceptor.ts — unified XHR/Fetch interceptor (was duplicated 3x)
- NEW: src/version.ts — centralized PKG_VERSION (was duplicated 2x)
- NEW: src/constants.ts — shared VOLATILE_PARAMS, FIELD_ROLES etc.
- NEW: src/engine.test.ts, src/registry.test.ts — 14 new unit tests

- browser.ts: use shared normalizeEval, interceptor, withTimeoutMs, PKG_VERSION
- intercept.ts, tap.ts: use shared interceptor generators
- cascade.ts: extract shared buildFetchProbeJs (90% dedup)
- engine.ts: use InternalCliCommand (no more 'as any' casts)
- executor.ts: remove all 15 'as StepHandler' type casts
- runtime.ts: add withTimeoutMs, IBrowserFactory interface
- registry.ts: add InternalCliCommand type for internal fields
- validate.ts: add pipeline step name validation
- output.ts: remove 'null as any' and '.filter(() => true)' hacks
- explore.ts, synthesize.ts: use shared constants
- docs: fix V2EX commands (3→6), SKILL.md version, verify example

Tests: 88 passed (was 74), tsc --noEmit: 0 errors
2026-03-16 13:22:12 +08:00
jackwener b0c51ddf19 fix: use Playwright MCP --executable-path flag (kebab-case)
Extract buildMcpArgs() helper and fix --executablePath → --executable-path
to match the Playwright MCP CLI's expected flag format.

Closes #16

Co-authored-by: KasumiChen <KasumiChen@users.noreply.github.com>
2026-03-16 12:58:18 +08:00
jackwener 141c2cf7d5 0.5.1 v0.5.1 2026-03-16 12:54:45 +08:00
jackwener 7ee0c9c96b fix: tab cleanup regex to match actual Playwright MCP tab format
The tab list from Playwright MCP uses '- N: (current) [title](url)' format,
but extractTabEntries only matched 'Tab N ...' format. This caused
_initialTabIdentities to always be empty, so tabs were never cleaned up.

Now supports both formats.
2026-03-16 12:54:44 +08:00
jackwener 76eefed83d docs: add opencli doctor hint after token setup 2026-03-16 12:47:59 +08:00
jackwener 08ae9c3fed chore: sync SKILL.md version to 0.5.0 2026-03-16 12:46:30 +08:00
jackwener 5a1581e67a 0.5.0 v0.5.0 2026-03-16 12:46:16 +08:00
jackwener 7a24e8d74b fix: add engines field for Node badge, sync SKILL.md version, fix zh-CN intro 2026-03-16 12:46:15 +08:00
jackwener 95fb841fc7 refactor: remove all CDP code, keep extension-only mode
- Remove CDP auto-discovery (discoverChromeEndpoint, isCdpApiAvailable, isPortReachable)
- Remove CDP readiness probe and autoAllowCdpDialog
- Remove forceExtension plumbing from registry, runtime, main, and adapters
- Remove CDP diagnostics from doctor (extractTokenViaCdp, remoteDebugging fields)
- Simplify connect() to always use --extension mode
- Remove CDP env vars (OPENCLI_USE_CDP, OPENCLI_CDP_ENDPOINT, OPENCLI_FORCE_EXTENSION)
- Update README.md, README.zh-CN.md, SKILL.md to remove CDP sections
- Update tests to remove CDP-related assertions

Current state preserved in extension-and-cdp branch.
2026-03-16 12:44:09 +08:00
jackwener 67d50191af fix: CDP mode hanging when using chrome://inspect remote debugging
- Add isCdpApiAvailable() to verify CDP HTTP JSON API before using endpoint
- Chrome's chrome://inspect#remote-debugging writes DevToolsActivePort but
  its CDP endpoint is incompatible with Playwright connectOverCDP (init
  succeeds but all tool calls hang silently)
- Add CDP readiness probe in connect() to catch unresponsive endpoints
- Skip tab cleanup in close() for CDP mode (no bridge tabs to clean)
- Add PlaywrightMCPMode tracking for extension vs CDP lifecycle policy
v0.4.6
2026-03-16 02:28:28 +08:00
jackwener 61a62f80c4 0.4.5 v0.4.5 2026-03-16 01:51:27 +08:00
jackwener f8491f01ae fix: add 5s timeout for initial tabs fetch during CDP connect
CDP mode on @playwright/mcp v0.0.68 hangs on browser_tabs calls.
This timeout prevents the entire connect() from blocking for 30s.
2026-03-16 01:51:22 +08:00
jackwener 4ec454530a 0.4.4 v0.4.4 2026-03-15 23:58:15 +08:00
jackwener e2192a2e8d docs: rename CLI-CREATOR → CLI-EXPLORER, add CLI-ONESHOT refs to READMEs
- Rename CLI-CREATOR.md → CLI-EXPLORER.md (explorer better reflects its purpose)
- Update all references across SKILL.md, README.md, README.zh-CN.md, CLI-ONESHOT.md
- Add quick-mode (CLI-ONESHOT) and full-mode (CLI-EXPLORER) entries in both READMEs
2026-03-15 23:58:10 +08:00
jackwener db8194337b Improve browser session lifecycle handling 2026-03-15 23:31:52 +08:00
jackwener 8e5a61458d docs: add CLI-ONESHOT.md — lightweight one-shot CLI generator guide
- CLI-ONESHOT.md: ~150 lines, 4-step flow for single-URL adapter generation
- Update SKILL.md: add quick-mode tip pointing to CLI-ONESHOT.md
- Update CLI-CREATOR.md: add tip pointing to lightweight alternative
2026-03-15 23:30:13 +08:00
jackwener dca908db90 docs: use bold AND to emphasize required config in zh-CN 2026-03-15 22:13:34 +08:00
jackwener 5a676b9f4b docs: clarify that both env var and MCP config are required for token 2026-03-15 22:12:59 +08:00
jackwener 797a392250 0.4.3 v0.4.3 2026-03-15 21:53:43 +08:00
jackwener 59a02d3a5d feat(twitter): implement like adapter 2026-03-15 21:53:42 +08:00
jackwener cef50dc8bf feat(doctor): print cli version in doctor report 2026-03-15 21:51:21 +08:00
jackwener 1aaafed2db feat(twitter): implement post, reply, delete, following, followers and notifications adapters 2026-03-15 21:42:48 +08:00
jackwener ee6669f45d feat(core): support YAML format output and optimize manifest builder 2026-03-15 21:05:36 +08:00
jackwener eb9050fadd feat(v2ex): add daily, me, and notifications session adapters 2026-03-15 20:59:39 +08:00
jackwener 53cd006a23 fix(boss): resolve limit=60 infinite loop and global CLI hangs
- search.ts: Added deduplication break condition so if the API returns duplicate pages, it stops instead of looping forever.
- search.ts: Removed unsafe Promise.race around page.goto which caused 'context closed' errors.
- browser.ts: Added robust static cleanup hooks to guarantee process and MCP lock destruction on crashed or ctrl+c'd runs.
2026-03-15 20:33:22 +08:00
jackwener c9710a656b feat(boss): add pagination loop to support fetching more than 15 results
- The API endpoint only returns 15 results per page.
- Added a loop to fetch multiple pages until the requested limit is reached.
- Added random 1-3 second delays between page fetches to avoid rate limits.
2026-03-15 20:19:30 +08:00
jackwener 4f570fa75e docs: add instruction to get token by clicking extension icon 2026-03-15 20:08:30 +08:00
jackwener 730285c0ca docs: set MCP Bridge as primary connection method, CDP as fallback 2026-03-15 20:07:32 +08:00
jackwener c5915d9ee3 docs: clarify that MCP token needs to be configured in both MCP config and env var 2026-03-15 20:06:32 +08:00
jackwener b7c8067bc9 docs: restore playwright MCP token configuration guide 2026-03-15 20:05:08 +08:00
jackwener c28ccd4b23 docs: huge improvement on README structure, badges and readability 2026-03-15 19:59:02 +08:00
jackwener 00f0ac9133 fix(boss): prioritize fresh browser cookie extraction over boss-cli cache 2026-03-15 19:58:46 +08:00
jackwener 64d655ae0d feat(boss): rewrite search adapter to use boss-cli native cookie extraction
- BOSS Zhipin aggressively detects CDP connections and kills pages
- Completely rewrote boss/search.ts to use browser: false (bypassing Playwright)
- Implemented native Python subprocess to extract cookies via browser-cookie3
- Falls back to boss-cli ~/.config/boss-cli/credential.json if available
- Uses direct Node.js fetch with anti-detection headers to avoid bans
2026-03-15 19:55:46 +08:00
jackwener 9c78ffdd39 chore: change license to BSD-3-Clause 2026-03-15 19:54:34 +08:00
jackwener a6e9c3b611 feat(browser): enhance CDP auto-discovery with port probe & opt-in priority
- Add TCP port reachability check to handle stale DevToolsActivePort files
- Make discoverChromeEndpoint() async with port probe before returning
- Change connection priority: extension mode is default, CDP is opt-in
  - OPENCLI_CDP_ENDPOINT: explicit CDP endpoint
  - OPENCLI_USE_CDP=1: enable DevToolsActivePort auto-discovery
  - Default: --extension mode (Playwright MCP Bridge)
- Add verbose logging for CDP discovery when -v flag is used
- Fix indentation inconsistency in connect() method
- Update README.md, README.zh-CN.md, SKILL.md with new connection docs
2026-03-15 19:44:27 +08:00
jackwener da155e96f0 feat: add forceExtension to bypass CDP detection for anti-bot sites
- Added forceExtension flag to CliCommand, CliOptions, browserSession, and PlaywrightMCP.connect()
- BOSS Zhipin detects CDP mode and drops connections; marked boss/search with forceExtension: true
- Also supports OPENCLI_FORCE_EXTENSION=1 env var for global override
- Fixed async discoverChromeEndpoint() placement outside Promise constructor
- Enhanced boss search with city names, experience, degree, salary, etc.
2026-03-15 19:39:39 +08:00
backtime1993 c20eaee393 feat(browser): Chrome 144+ auto-discovery via DevToolsActivePort
Chrome 144+ allows enabling remote debugging from chrome://inspect
without any command-line flags (--remote-debugging-port). Chrome writes
the active port and browser GUID to a DevToolsActivePort file in the
user data directory.

This change reads that file to auto-discover the CDP WebSocket endpoint,
eliminating the need for the Playwright MCP Bridge browser extension in
most cases.

Connection priority:
1. OPENCLI_CDP_ENDPOINT env var (manual override)
2. DevToolsActivePort auto-discovery (Chrome/Edge, cross-platform)
3. --extension mode fallback (original behavior)

Supports Windows, macOS, and Linux with Chrome, Edge, and Chromium.

Co-authored-by: kensei <backtime1993@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-15 19:34:14 +08:00
jackwener 15d8c45b95 feat(boss): add city name mapping, experience/degree/salary filters, page support
- 30+ Chinese city names → code mapping (e.g. --city 杭州)
- Experience filter: 应届/1-3年/3-5年/5-10年/10年以上
- Degree filter: 大专/本科/硕士/博士
- Salary filter: 3K以下/3-5K/5-10K/10-20K/20-50K/50K以上
- Page number support for pagination
- Area now shows city·district·business district
- Skills column added to output
2026-03-15 19:31:05 +08:00
jackwener 3ac5aa3023 docs: reduce excessive emoji usage across markdown files 2026-03-15 19:30:32 +08:00
jackwener 9a6e157af6 docs: add strong AI Agent prompts to read CLI-CREATOR.md before creating adapters 2026-03-15 19:28:05 +08:00
jackwener 9a45988197 v0.4.2 v0.4.2 2026-03-15 19:00:15 +08:00
jackwener 2d638e3a0a fix(manifest): extract TS adapter args at build time for proper CLI registration
The manifest builder now statically parses TS source files to extract
args, description, columns, domain, and strategy metadata. Previously
TS adapters had empty args in the manifest, causing commander to reject
their options (e.g. --keyword) before the lazy module could load.

Also includes the xiaohongshu search DOM fallback fix from earlier.

Co-authored-by: jack4world <jack4world@users.noreply.github.com>
2026-03-15 18:59:15 +08:00
jackwener b3850d9c53 fix(xiaohongshu): switch search from broken XHR interception to DOM extraction (Fixes #10)
Co-authored-by: jack4world <jack4world@users.noreply.github.com>
2026-03-15 18:55:29 +08:00
jackwener 91d7892516 docs(CLI-CREATOR): streamline 840→719 lines — remove redundancies, keep all value
- MCP Bridge: 6-step flowchart + step-by-step → compact table (-51 lines)
- Remove duplicate cascade code example (kept full template below)
- Verbose mode + output verification → 3-line compact form
- Step 5 + design philosophy → merged into concise section
2026-03-15 18:49:47 +08:00
jackwener 3f859b3238 docs(CLI-CREATOR): add SDK table, template-first guide, pagination pattern, success story
Four practical improvements based on real adapter development experience:
- Real success story: 5-minute following adapter walkthrough
- Step 2.5: template-first + Bilibili SDK reference table
- Pagination pattern with code template (pn+ps)
- Wbi vs fetchJson decision rule
2026-03-15 18:46:13 +08:00
jackwener 2fdab085a7 chore: release v0.4.1 v0.4.1 2026-03-15 18:41:14 +08:00
jackwener 7ec04c66cb docs: deep review fixes — sync command tables to actual 47 commands
- Command count 46 → 47 (added following)
- bilibili row: add 'following' command
- github row: remove non-existent 'trending', change mode to Public
- SKILL.md: add following, xueqiu hot/search commands
- Fix github section from 'trending=browser' to 'public'
2026-03-15 18:41:14 +08:00
jackwener f8ee50b776 feat(bilibili): add following adapter — get user's followed accounts
Discovered via Playwright MCP Bridge workflow (CLI-CREATOR.md):
 - Navigate to space.bilibili.com/{uid}/fans/follow
 - Sniff network: GET /x/relation/followings (Cookie-tier, no Wbi needed)
 - Verify API response: total=1342, fields: mid/uname/sign

Also fixed .d.ts loading bug in build-manifest.ts and engine.ts:
 - build-manifest now scans .js files in dist/ (not .ts)
 - engine.ts skips .d.js declaration stubs
2026-03-15 18:38:34 +08:00