feat(job): add get_saved_jobs tool

[alfonzeta]

Summary

• Adds get_saved_jobs MCP tool to list job postings saved by the authenticated user at /my-items/saved-jobs/
• Returns {url, sections: {saved_jobs}, job_ids, references?} compatible with existing get_job_details
• Supports pagination via max_pages (1–10, default 3) using ?start= offsets
• Locale-independent extraction: innerText + a[href*="/jobs/view/"] job ID parsing

Closes #522

Synthetic prompt

Add get_saved_jobs tool following the search_jobs pattern: navigate to /my-items/saved-jobs/, extract innerText and job IDs with pagination, register in tools/job.py, add tests and update README, docker-hub.md, and manifest.json.

Test plan

• uv run pytest tests/test_scraping.py::TestGetSavedJobs
• uv run pytest tests/test_tools.py::TestJobTools::test_get_saved_jobs
• Live verification against authenticated session (get_saved_jobs(max_pages=2) returned 20 job IDs across 2 pages)
• Maintainer review on non-English LinkedIn locale (extraction is locale-independent by design)

[greptile-apps]

Greptile Summary

This PR adds a saved-jobs tool for LinkedIn job workflows. The main changes are:

• New get_saved_jobs MCP tool registration.
• Saved-jobs scraping and job ID extraction from /my-items/saved-jobs/.
• Pagination, diagnostics, docs, manifest, and tests for the new tool.

Confidence Score: 4/5

This is close, but the saved-jobs pagination should be fixed before merging.

• The new tool can return incomplete saved-job results for users with more than one saved-jobs page.
• The wrapper, diagnostics, docs, and manifest changes look consistent with the existing job tools.

linkedin_mcp_server/scraping/extractor.py

Important Files Changed

Filename	Overview
linkedin_mcp_server/scraping/extractor.py	Adds the saved-jobs extractor and pagination logic, including the offset issue that can skip saved jobs.
linkedin_mcp_server/tools/job.py	Registers the new MCP tool wrapper and forwards max_pages to the extractor.

Prompt To Fix All With AI

Fix the following 1 code review issue. Work through them one at a time, proposing concise fixes.

---

### Issue 1 of 1
linkedin_mcp_server/scraping/extractor.py:3376-3380
**Use saved-jobs offset** The saved-jobs path reuses `_PAGE_SIZE = 25` from job search, but the PR's live check says two saved-job pages returned 20 IDs, so this list paginates in 10-item pages. With `max_pages=2`, the second request jumps to `?start=25`, skipping saved jobs at offsets 10 through 24 and returning an incomplete list for anyone with more than one page of saved jobs.

```suggestion
            saved_jobs_page_size = 10
            url = (
                base_url
                if page_num == 0
                else f"{base_url}?start={page_num * saved_jobs_page_size}"
            )
```

_{Reviews (1): Last reviewed commit: "style: ruff format saved jobs warning lo..." | Re-trigger Greptile}

[greptile-apps]

Use saved-jobs offset The saved-jobs path reuses _PAGE_SIZE = 25 from job search, but the PR's live check says two saved-job pages returned 20 IDs, so this list paginates in 10-item pages. With max_pages=2, the second request jumps to ?start=25, skipping saved jobs at offsets 10 through 24 and returning an incomplete list for anyone with more than one page of saved jobs.

Suggested change

	url = (
	base_url
	if page_num == 0
	else f"{base_url}?start={page_num * _PAGE_SIZE}"
	)
	saved_jobs_page_size = 10
	url = (
	base_url
	if page_num == 0
	else f"{base_url}?start={page_num * saved_jobs_page_size}"
	)

Prompt To Fix With AI

This is a comment left during a code review.
Path: linkedin_mcp_server/scraping/extractor.py
Line: 3376-3380

Comment:
**Use saved-jobs offset** The saved-jobs path reuses `_PAGE_SIZE = 25` from job search, but the PR's live check says two saved-job pages returned 20 IDs, so this list paginates in 10-item pages. With `max_pages=2`, the second request jumps to `?start=25`, skipping saved jobs at offsets 10 through 24 and returning an incomplete list for anyone with more than one page of saved jobs.

```suggestion
            saved_jobs_page_size = 10
            url = (
                base_url
                if page_num == 0
                else f"{base_url}?start={page_num * saved_jobs_page_size}"
            )
```

How can I resolve this? If you propose a fix, please make it concise.