Stop refreshing forty job boards, and stop re-explaining yourself to a chatbot every session.
TekJobs turns your resume, your preferences, your decisions and your application history into a career record on your own machine that your AI works from. Every morning it scans hundreds of public company boards, says why each role matched, drafts truthful application material, and stops before anything is sent.
The model can change. Your context stays with you.
The morning scan, the cut to the few that fit, every point written down, the Today list, the pipeline, your AI over MCP, and the one step it leaves to you.
Motion graphics, not a screen recording. The companies are the fictional ones in the sample vault; the reasons, the scores and the tool names are what TekJobs itself writes and serves. The postings count is illustrative. For one real, unedited run, watch the demo below. Sharing it? There is a vertical cut and a square cut.
Fifty-seven seconds, recorded from one real run on the fictional sample resume: the init command, the interview in Claude Code, the first scan of 304 boards, and the Today page it produced.
Nothing in it is staged: the terminal output, the interview transcript and the screenshots are from the same run, and only the folder path was shortened. The companies on the Today page are real postings found that morning.
A scheduled task does the first three before you wake up. The last two are yours, and the app keeps them short.
Every morning: the 304 company boards in the starter registry (Greenhouse, Lever, Ashby, Workday and the rest), the Google and Apple career pages, a dozen aggregator feeds, and any boards you add. Tens of thousands of postings a run, deduplicated. All public; the company-board scan needs no API key, and the two optional feeds that do (Adzuna, USAJOBS) bring their own free ones.
Each posting is scored against criteria you own: title terms, hard exclusions, description keywords, seniority, a remote rule, a pay floor with a stretch band, recency. Every point is written down as a reason.
A match becomes a markdown note in a folder on your machine: the posting, why it matched, a status log, an application packet. Open the folder in Obsidian or anything else. The notes are the record; nothing else is.
Today shows the few worth a decision and what is about to close. Shortlist or pass, with a reason. The mail pass reads your inbox for confirmations, rejections and interview invitations, then asks you to confirm each change before a note moves.
A tailored resume and a letter written from your resume of record, checked against it, saved into the note. The copy panel holds the answers forms keep asking for. You review, you click send; the app never does.
Dark by default, one accent used sparingly, every number in mono. Filters are chips, views are links, and the sheet says why a job scored what it scored.
Most AI job-search tools begin with an empty prompt. TekJobs begins with what you already know: your experience, what you want, what you turned down and why, where you applied. Your AI works from that record instead of rediscovering you every session, and the record gets better the longer you search.
A job search is a series of claims about things that happened to you. The software drafts and files; the claims stay yours.
Drafts are written into the note and stop there. Applied, interviewing and offer are statuses only a person can set, because only a person knows.
No profile scraping, no automated messages. A single job link you paste is read once, the way a browser would open it.
The mail check runs with only Gmail's read tools allowed and every write tool denied by name. The model extracts; matching and every change to a note is ordinary code, and nothing changes until you confirm.
The scan runs on your machine, the notes live in your folder, and there is no TekJobs account or server holding a copy. What does leave: the requests to public job boards, and whatever context you choose to send to the AI provider you connect, which processes it under its own terms. No telemetry, no separate model API key, nothing sent on your behalf.
TekJobs users do not browse a board. Their software reads hundreds of them every morning and scores each posting against criteria they wrote. A posting here is one more source in that scan: no inbox blast, no promoted slot, no list of who looked. The people it fits find it, with the reasons.
Title, company, location and whether it is remote, the pay range, and the description. The pay range is required: a posting without one scores lower in every TekJobs user's criteria, and most of them have a floor.
$49 for 30 days, through Stripe. No subscription, no per-view charge, no upsell. The refund policy is one page.
The posting becomes one more source every TekJobs user's software reads the next morning, scored against that user's own criteria like any Greenhouse or Ashby board. It is never emailed, pushed or promoted; it is found.
A user whose criteria the posting clears sees it on their Today page with the reasons it scored, and applies through your own link. We are not in the middle of the application, and we never hand you a list of who saw it.
One posting, paid once. It runs for 30 days or until you close it.
The board and tool counts are read from the repository on every build. The postings figure is from the author's own vault on 22 September 2026. The board registry is in the repository as a markdown table, and pull requests to it are the most useful contribution.
In the starter registry every new profile folder begins with: eleven board platforms plus the Atlassian, GitHub, Spotify and Amazon career APIs and the Google and Apple pages. Add your own; the author's vault watches 356.
The author's vault, 22 September 2026: fetched, deduplicated across sources, scored, and cut to the ones that clear the bar. About a minute on a laptop.
Claude Code, Codex, Cursor or Claude Desktop can run the whole search over the same notes: find, move, draft, add boards, start scans, read the mailbox.
No TekJobs account, no hosted search database, no separate model API key. A folder of markdown, a scheduled task, and the AI client you already have.
Nothing personal lives in the repository. Your profile folder, criteria, notes and mail state are outside it; the repository is the code and the starter notes.
One command creates your profile folder and imports the resume. The interview then runs inside the AI client you already have: it reads the resume, asks the few things a resume cannot say, writes your profile and criteria, and runs the first scan. After that a scheduled task does the mornings.
About twenty-five minutes end to end on a fresh machine, most of it the interview. Not ready to use your own resume? The repository ships a fictional profile folder to look around in first.
Everything after the first scan is in the docs: the profile folder, scoring, mail, people, the copy panel, letters and resumes, and the 44 MCP tools.
npx @timurtekb/tekjobs init ~/Obsidian/JobSearch --resume ~/Downloads/resume.pdf
npm install -g @timurtekb/tekjobs # so the morning task has a command that lasts
# Connect your AI client. Claude Code is one line; Codex, Cursor and Claude Desktop are in the docs.
claude mcp add tekjobs -- tekjobs mcp
# Then say:
# "Use the tekjobs MCP server. Call onboarding_status, then onboarding_materials, and follow its script."
tekjobs serve # the app on http://127.0.0.1:8787
tekjobs schedule # the morning scan, every day at 07:30