GigWatch

A self-hosted freelance-gig watcher. GigWatch monitors live job/gig feeds, filters them against your skills, remembers what it has already shown you, and alerts you (console, email, or Slack) only when a new matching gig appears.

Stop refreshing Upwork/Remotive/LinkedIn every 20 minutes. Point GigWatch at the feeds you care about, tell it what you do, and let it ping you when something relevant lands.

$ gigwatch scan
scanned 17 job(s) from 1 source(s); 4 match filter; 2 new
GigWatch: 2 new matching gig(s)

1. Senior Python Backend Engineer
   company: Acme Digital
   salary:  $120k-$150k
   where:   Remote (Worldwide)
   matched: python, backend
   score:   9.0
   https://remotive.com/remote-jobs/...

Why this exists

Freelancers lose real money to latency — the best gigs get filled in the first hours. Existing monitors (Distill, PageCrawl, Apify's Upwork monitor) are hosted SaaS that scrape your sessions and cost monthly. GigWatch is:

• Self-hosted & private — your skills, your feeds, your machine. No account, no session scraping, no data leaving your box.
• Free & open (MIT) — the core is a small, readable Python CLI.
• Portable — run it on a laptop, a $5 VPS, or in a cron job.

Features

• Multiple feed sources — Remotive, We Work Remotely, and RemoteOK (built-in, no auth), any RSS/Atom feed, or any JSON endpoint returning a list of job objects.
• Output formats — scan and list take --format text|markdown|json (a Markdown table or a JSON array of job objects for piping elsewhere).
• Skill-based filtering — keyword matching (any/all), category and location filters, exclude-list, and a relevance score (title hits weigh more than body hits).
• Dedupe by state — a local JSON state file tracks what you've already seen, so you're only ever alerted on new matches. State auto-prunes.
• Alerts — console (default), SMTP email, or Slack (webhook or chat API). Secrets come from environment variables, never the config file.
• Zero dependencies — pure Python standard library. If you can run python3, you can run GigWatch.

Quick start

No install required — it's stdlib-only:

# 1. Get the code
git clone https://github.com/earnnova7/gigwatch
cd gigwatch

# 2. Create a starter config (or copy config.example.json to config.json)
python3 -m gigwatch init

# 3. Edit config.json: put YOUR skills in filters.keywords
#    e.g. ["python", "backend", "api", "django"]

# 4. See what would match right now (dry run, no state touched)
python3 -m gigwatch list

# 5. Do a real scan: alerts on new matches, remembers them
python3 -m gigwatch scan

Or install it as a proper command:

pip install .          # or: pipx install .
gigwatch init && gigwatch scan

Run it continuously

# Loop forever, re-scanning every poll_interval seconds (default 15 min):
gigwatch watch

# Or use cron on a VPS (once an hour):
0 * * * * cd /opt/gigwatch && /usr/bin/python3 -m gigwatch scan

Alerts

Console is on by default. To also get email/Slack, fill in the alerts section of config.json and set the env vars it references:

export GIGWATCH_EMAIL_TO=you@yourdomain.com
export GIGWATCH_SMTP_USER=...
export GIGWATCH_SMTP_PASS=...
export GIGWATCH_SLACK_WEBHOOK=https://hooks.slack.com/services/...

The ${VAR} placeholders in the config are expanded from the environment at load time, so secrets never live in the file (and the file is safe to commit if you want).

Configuration

See config.example.json for a fully annotated example. The top-level keys:

Adding your own JSON source

Any endpoint that returns a JSON array of objects works. Each object should have at least title and url; company/company_name, category, location/country, salary, tags, and description are picked up if present. Wrap in {"jobs":[...]}, {"data":[...]}, {"results":[...]}, or {"items":[...]} and it still works.

Commands

Useful flags: --config PATH (default config.json), -v/--verbose, --max-age-days N (state pruning; 0 keeps everything), and --format text|markdown|json on scan/list.

How it works

sources ──fetch──> [Job, Job, ...]
                        │
                   filter_jobs()  (keywords / category / location / score)
                        │
                   [ScoredJob, ...]
                        │
              compare against state file
                        │
                 ┌──────┴──────┐
              new jobs       already seen
                 │               │
           alert (console/     drop silently
           email/slack)
                 │
         mark seen in state file  ──>  gigwatch-state.json

The state file is plain JSON ({job_id: first_seen_utc}) so you can inspect it, back it up, or move it between machines.

Roadmap / ideas

• More built-in sources (Hacker News "Who is hiring", LinkedIn via RSS, Upwork via a user-supplied export).
• AI ranking: summarize each match and rank by fit to a profile you write.
• Draft proposals / cover letters per match.
• A tiny hosted tier (the natural monetization path — see below).

Monetization

The core is free and MIT-licensed. The obvious paid tier is a hosted GigWatch: we run the scanner on our infra, you get email/Slack alerts without running anything, plus AI-ranked matches and proposal drafts. That's the plan — the open-source CLI is both the product's engine and the marketing.

License

MIT — see LICENSE.

Contributing

Issues and PRs welcome. The code is deliberately small and stdlib-only; keep it that way. Run pytest (unit) and pytest -m live (hits the real Remotive API) before sending a PR.