bike-theft-shield
Use when a bike might be stolen (locking up in a city, buying a lock) or HAS been stolen. Grades lock setups against theft-area risk, builds a serial/photo registry BEFORE theft, and prints the first-72-hours stolen-bike playbook: police report, registry flags, marketplace sting searches, insurance claim.
Bike Theft Shield
Overview
Roughly 2 million bikes are stolen every year in North America alone, and the recovery rate is dismal (<3% for unregistered bikes) — not because bikes vanish, but because owners can't produce a serial number in the first hour, and stolen bikes are resold within 48 hours on marketplace apps. Prevention and recovery are both systems, not products: a lock graded for the parking location, a registry entry that survives the theft, and a pre-built playbook that fires in the first 72 hours while the bike is still local.
This skill has three modes: harden (lock audit + locking protocol + registration), register (serial + photo evidence file, stored locally as JSON), and stolen (the 72-hour recovery playbook with report templates and sting-search setup).
When to Use
- Moving to / commuting in a city, or parking a bike outside overnight anywhere
- Buying a lock or judging whether the current lock survives the parking spot
- Buying a used bike (check it isn't stolen — serial + registry check)
- A bike was just stolen — run
stolenimmediately - Don't use for: motorbike/scooter theft (different registries and locks), inside-home storage audits only.
Quick Start
# 1. Is my lock good enough for where I park?
python3 scripts/bikeshield.py audit --bike-value 1400 --area-risk 4 \
--locks "kryptonite evolution u-lock; cable"
# 2. Register the bike BEFORE anything happens (serial + photo checklist)
python3 scripts/bikeshield.py register add --make "Trek" --model "Marlin 7" \
--year 2023 --color "Black" --value 1400 --serial "WTU123A4567"
python3 scripts/bikeshield.py register photos trek-marlin # what to photograph
# 3. THEFT: first 72 hours playbook
python3 scripts/bikeshield.py stolen --bike-file bikes.json --hours-ago 2
# 4. Used-bike purchase check
python3 scripts/bikeshield.py check --serial "WTU123A4567"
Lock Grading Model
| Rating (Sold Secure / ART) | Resists | Fails to | Suitable area risk |
|---|---|---|---|
| Diamond | 60s+ angle grinder attack, all hand attacks | sustained cordless grinder x2 | 4–5 (city cores, overnight) |
| Gold (≈ART 4) | bolt croppers to 42", leverage, hammers | 1 grinder ~1–2 min | 3–4 (commuter parking, daytime) |
| Silver (≈ART 2–3) | basic hand attacks, small croppers | big croppers, grinders | 2 (campus, short stops) |
| Bronze / cable / combo | opportunistic tugs only | literally everything else | 1 (in-sight only) |
Spending doctrine: lock budget ≈ 10–15% of bike value in risk 3+, always as a pair (primary U/chain + secondary for wheels), because a single lock leaves a quick-release wheel or seatpost as the cheap win.
Locking Protocol (what the audit prints)
- Object test: street signs (unbolt-able), scaffolding, chain-link, trees = NO; ground-anchored rack, bike hoop, fence post capped = YES.
- Sheldon method: U-lock around rear wheel + seat tube inside the rear triangle — wheel+frame secured with a small lock. Fill the shackle space (leave no room for a jack).
- Quick releases: replace skewers with locking ones or take the seat/light with you.
- Lock high (off the ground = no bottle-jack leverage), keyway facing down.
- Never lock overnight outside in risk 4–5 — no lock survives a night grinder attack.
Stolen-Bike 72-Hour Playbook (summary)
- Hour 0–1: file the police report (a case number unlocks everything else: insurance, registry flag, marketplace takedowns). Flag stolen on BikeIndex + Project 529 + national registry if applicable.
- Hour 1–24: create saved searches on Facebook Marketplace, Craigslist, OfferUp/eBay with the model name + city + 40 km radius. Stolen bikes surface locally within 48 h. Set alerts for "bike" + your frame size as well (thieves strip details).
- Day 1–3: print QR flyers for bike shops, pawn shops, co-ops (they all check serials); post in local cycling Facebook/Reddit/WhatsApp groups with the registry link — these communities recover more bikes than police do.
- Recovery sting: never meet alone; agree on inspecting "your" bike by serial under a pretext (buying), bring the police case number and one friend; if serial matches, call the non-emergency line instead of confronting.
- Insurance: homeowners/renters covers bikes often with a sublimit (~$500–1500) — file within the policy window; check if the policy needs the police report within 24–72 h.
Common Pitfalls
- Cable lock as the only lock. Every cable falls to 24" bolt croppers in seconds. Cables are wheel-accessories, never primary.
- Serial number unknown until after the theft. It's stamped on the frame (bottom bracket shell mostly — per-brand locations in the script); the police report needs it and marketplace takedowns require it.
- Locking the top tube to a sign post. Lift the bike over the post — gone. Object test first, always.
- Photos of the drive side only. Registry photos: both sides, drivetrain close-up, serial close-up, distinctive scratches, components (wheelset, saddle) — these identify stripped frames.
- Meeting a "seller" who has your bike. Confrontations go badly; use the serial + police case number + non-emergency line, or bring a uniformed officer to the meeting (many departments do this gladly).
Verification Checklist
- Audit verdict names the weakest link (lock, wheel, or object) — not just a grade
- Registry JSON has: serial, 6+ photos checklist, distinct marks, purchase proof
- Serial location guidance matches the bike's brand
- Stolen playbook step 1 = police report with case number before anything else
- Insurance sublimit and claim deadline checked against the policy
Example Session
$ python3 scripts/bikeshield.py audit --bike-value 1400 --area-risk 4 \
--locks "kryptonite evolution u-lock; cable"
LOCK AUDIT — bike $1400, area risk 4/5
══════════════════════════════════════
[1] kryptonite evolution u-lock → Gold (≈ART 4)
Resists croppers/leverage; falls to a cordless grinder in 1–2 min.
OK for daytime commuter parking; NOT for overnight in risk 4.
[2] cable → NOT A LOCK (opportunistic-tug class only)
24" bolt croppers cut it in seconds — demote to wheel duty.
VERDICT: PARTIAL — frame protected by day logic only.
WEAKEST LINK: front wheel (quick-release) + overnight exposure.
FIX: • Add locking skewers or a secondary Gold mini-U for the front wheel.
• Lock budget check: $140–210 recommended — currently ~$80 → underinvested.
• Overnight rule: bike goes inside; no lock survives a night grinder run.
- Downloads
- 89All time
- Security audit
- Pass
- Last updated
- 3d ago
- Current version
- v1.0.0
- License
- MIT-0
