Wolverine

HEAL. FEED.

TRUST.

Restock radar that will not cry wolf. Studio scrapes niche electronics; Heal Court decides release, repair, or refuse.

// SECTION: SCAR_FEED + HEAL_COURT
004
heal_loop.sys
_
store_radar.dither320x240
scar.metrics
0Studio Collectors
00+Trusted Signals
0Court Verdicts
0.0KCatalog Rows
heal_court.statusTICK:0000
StoreVerdictBatch
ADAFRUIT
RELEASE
ok
SPARKFUN
RELEASE
ok
PIMORONI
RELEASE
ok
THEPIHUT
RELEASE
ok
Signal Trust100%
// SECTION: ABOUT_WOLVERINE
005
RENDER: field_pipeline.objLIVE
Isometric field infrastructure for Scar Feed pipeline
CAM: -45deg / ISORES: 2048x2048
MANIFEST.mdv0.1.7

Broken scrapers invent restocks.
We refuse to cry wolf.

Makers hunt scarce boards across Adafruit, SparkFun, Pimoroni, and The Pi Hut. A clone price or empty stock field looks like “back in stock” and wastes a buy. Judges care about self-heal — we made it visible.

Wolverine is agent-driven Scraper Studio: create, run, heal, approve. Heal Court verdicts gate every signal. Scar Feed only speaks when the batch is trusted.

FIELD_UPTIME:0d 00h 00m 00s
COLLECTORS4
STORES4
VERDICTS3
TRUSTED60+
// SECTION: SCRAPER_STUDIO
006

Best Use of Bright Data — pinned collectors, agent CLI, real heals

Not a prebuilt dataset. Four Studio collectors for Adafruit, SparkFun, Pimoroni, and The Pi Hut — driven by coding agent + bdata, written to SQLite, gated by Heal Court, exported into Scar Feed.

collectors.pinNEVER RECREATE
  • Adafruitc_msyvm0ar1gznj2dlrq
  • SparkFunc_msywbl7b18fsthmxn
  • Pimoronic_msywj65f19rulm4cua
  • The Pi Hutc_msyx5sb61lfwvvvspd
agent.cliLOOP
  • bdata scraper createCustom collectors for long-tail stores
  • bdata scraper runStructured JSON → SQLite snapshots
  • bdata scraper healSelf-heal when selectors break
  • bdata scraper approveConfirm fix, re-run, prove it
// SECTION: JUDGE_EVIDENCE
007

Evidence over a staged feed dump

Git history, heal journal, Actions cron, and sample structured output. The demo is the product — public UI ships scar.json, never your Bright Data key.

SOURCE · github.com/manasdutta04/wolverine-scraperStar on GitHub →
// SECTION: RUN_IT
008
docker.hubIMAGE

Run locally with Docker

Pull the dashboard image and open localhost:3000 — same Scar Feed workspace as this site.

Full docs →