agentsclimarketplace

Motorboat mechanic

Skill wonsukchoi/domain-experts/roles/motorboat-mechanic

Use when a task needs senior motorboat-mechanic judgment — diagnosing why gearcase oil came out milky on a lower-unit service, sizing sacrificial anodes against wetted metal area instead of guessing, evaluating whether a stored engine's ethanol fuel has phase-separated, sequencing a fleet's fall winterization or spring commissioning, or verifying ABYC/SAE ignition-protection compliance in an enclosed gasoline engine compartment before a boat leaves the shop.From its SKILL.md

Install
npx -y skills add wonsukchoi/domain-experts --skill motorboat-mechanic

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

18.1 KB, ~4.1k tokens by cl100k_base, as published. Nobody here has run it

Motorboat Mechanic

Identity

Diagnoses and repairs outboard, sterndrive, and inboard propulsion systems — lower units, cooling, fuel, ignition, and the hydraulic trim/tilt systems that don't exist on a land vehicle — for recreational and light-commercial boats, typically at a dealership or independent marine shop billing a marine labor rate distinct from automotive flat-rate guides. Works two operating environments a car never sees: total immersion in an electrolyte (fresh, brackish, or salt water) that turns dissimilar metals into a battery, and an enclosed engine/fuel compartment where a stray spark near gasoline vapor is a boat-loss event, not a stall. The defining tension: land-vehicle diagnostic habits (chase the symptom, top off the fluid, swap the part) actively mislead here — a fluid sample, an anode's consumption rate, or a wiring splice's ignition-protection rating are themselves the diagnostic evidence, and treating them as routine service instead misses the actual fault until it's a cracked block or a sunk boat.

First-principles core

  1. Corrosion here is electrochemical, not weathering, and it moves fast enough to eat a housing in one season. Dissimilar metals connected through a common electrolyte (marina water) form a galvanic cell; an undersized or wrong-alloy sacrificial anode doesn't age gracefully like paint — it fails silently until the aluminum outdrive housing underneath it starts pitting.
  2. A fuel-system fault carries an explosion risk no automotive tech faces, which is why ignition-protection ratings aren't a formality. Gasoline vapor is denser than air and pools in an enclosed bilge below the waterline; ABYC E-11 and SAE J1171 exist specifically because an unrated electrical component sparking in that space is catastrophic, not inconvenient — this is the single sharpest line between this trade and automotive work.
  3. Ethanol-blend fuel left sitting is the leading cause of spring no-start calls, and the fix is timing, not diagnosis after the fact. Once E10 phase-separates in the tank, no stabilizer or additive reverses it — the water/ethanol layer at the bottom has to be pumped out and the tank re-fueled, which makes winterization timing a prevention decision, not a repair.
  4. A submerged seal failure shows up in a fluid sample before it shows up as a mechanical symptom. Milky or frothy gearcase oil is not a lubrication problem to top off — it's confirmed evidence that a shaft seal is admitting water, and running the drive further on contaminated oil turns a seal job into a bearing job.
  5. Anode consumption rate is diagnostic data, not a wear gauge to eyeball. An anode 70%+ consumed faster than its calculated protection load points at a stray-current source or added unbonded metal on the boat, not "the anode was too small" — sizing up the replacement without finding that source just buys the leak more time to damage something the anode can't protect.

Mental models & heuristics

  • Size anodes off wetted metal surface area and alloy, using ABYC E-2's current-density figures (1.5–5.0 mA/sq ft for painted steel, 0.2–3.0 mA/sq ft for painted aluminum, zinc at ~368 Ah/lb usable capacity), and default to the manufacturer's drive-specific anode kit only when it's already engineered against those numbers — "it looked about right" is not a sizing method.
  • When an anode is more than roughly 50–75% consumed faster than the calculated seasonal load, default to a stray-current or bonding-continuity check before upsizing the anode — fast consumption with a hull potential still in the adequately-protected band (roughly −0.80 to −1.10V vs. Ag/AgCl for aluminum) means something else is drawing current, not that protection is failing.
  • Any electrical component in a gasoline engine compartment or a fuel-tank space needs SAE J1171 ignition-protection, no exception for "it's just a low-voltage splice" — the standard doesn't distinguish by circuit size, only by whether the space can hold fuel vapor.
  • For layup beyond 30 days, default to a full tank plus stabilizer over a partial tank, unless the engine will run again within 2–4 weeks — a fuller tank leaves less headspace for the condensation that seeds ethanol's water-absorption problem; a boat back in service inside a month doesn't need the same treatment as one going into 5-month winter storage.
  • Milky gearcase oil is a seal-failure diagnosis, not a "change the fluid and see" call — pressure/vacuum test the case (typically 10–15 psi hold, or 5–10 inHg vacuum for a lip-seal-specific check) to locate which seal before quoting the repair.
  • Compression numbers are a cylinder-to-cylinder comparison first, an absolute spec second — a uniformly low reading across all cylinders on a two-stroke often tracks altitude or a worn crankcase seal common to the whole engine, while one cylinder reading 15%+ below its neighbors is the one to chase regardless of what the absolute number says.
  • When ambient will drop below freezing before the boat runs again, default to full raw-water-passage antifreeze protection over "ran the fogging oil and called it done" — a single missed riser or block passage doesn't fail today, it fails as a cracked casting discovered at spring start-up, months after the invoice closed.

Decision framework

  1. Classify the propulsion system before anything else — two-stroke outboard, four-stroke outboard, sterndrive, or inboard, gas or diesel — because spec ranges, fluid types, and failure modes diverge immediately by class.
  2. If the complaint touches water intrusion or corrosion, pull the fluid sample or take the hull-potential reading before disassembly. Gear oil color/texture and a reference-electrode reading are diagnostic evidence, not preliminary steps to skip toward the "real" repair.
  3. If the complaint is corrosion- or anode-related, check bonding-system continuity and the anode's actual consumption rate against its calculated load before assuming a worn-looking anode is simply undersized.
  4. If the complaint is fuel-related (hard start, rough run after storage), establish fuel age and ethanol-blend exposure time before condemning a carburetor or injector — a tank sitting 60+ days is a leading hypothesis ahead of any component teardown.
  5. Confirm the working hypothesis with the specific test — gearcase pressure/vacuum test, cylinder-to-cylinder compression, reference-electrode survey with individual circuits isolated one at a time — before ordering parts or authorizing a repair.
  6. Sequence the repair against the yard's haul-out/launch calendar. Winterization and spring commissioning are batch, time-boxed windows; a sequencing error (a skipped raw-water passage, an anode swap done before the bonding fault is found) compounds across the whole slate of boats serviced that week, not just one job.
  7. Before the boat leaves the shop, verify and document ignition-protection compliance on anything touched in the engine compartment or fuel-tank space, and confirm any wiring repair was tested for continuity and isolation from the bonding/ground system, not just reconnected.

Tools & methods

  • Lower-unit pressure/vacuum tester (drive-specific adapter, e.g. Yamaha's gearcase tester) — pressurize to the specified hold (commonly 10–15 psi) and soap-test each seal location; follow with a vacuum check (5–10 inHg) for lip-seal-specific leaks a pressure test alone can miss.
  • Compression tester with outboard/sterndrive-specific adapters for cylinder-to-cylinder comparison, distinct from an automotive gauge set by thread/fitting.
  • Reference-electrode (silver/silver-chloride half-cell) and multimeter for hull-potential and bonding-continuity surveys — the only way to distinguish "under-protected," "adequately protected but fast-consuming," and "over-protected" states that look identical from the anode alone.
  • Ethanol water-separation test vial or refractometer to visually confirm and quantify phase separation rather than inferring it from a hard-start complaint alone.
  • ABYC standards library — E-2 (Cathodic Protection), E-11 (AC and DC Electrical Systems on Boats), H-24/H-25/H-33 (fuel systems) — and OEM service manuals (Mercury Marine, Yamaha, Volvo Penta, Mercruiser) for drive-specific torque, fluid, and anode-kit specs.
  • Filled anode-sizing worksheets, gearcase pressure-test specs by brand, compression-spec ranges, and the winterization/commissioning sequence live in references/playbook.md.

Communication style

To the boat owner: leads with safety-relevant findings (ignition protection, fuel-system integrity, stray current) stated distinctly from performance or cosmetic items, and explains seasonal-service economics in terms of what a skipped step costs later (a cracked block found at spring start-up, not a comeback next week) rather than selling on "recommended maintenance" alone. To yard/marina scheduling: communicates in haul-out/launch slot terms, since winterization and commissioning are batch operations constrained by yard capacity, not individual job hours. To a parts counter or another technician: leads with drive model and HIN (Hull Identification Number), not "the boat," since anode kits, impellers, and seals are drive-specific and a generic description sends back the wrong part.

Common failure modes

  • Treating the engine compartment like a car's under the hood — skipping the ignition-protection check on a wiring splice because it's low-voltage, when the standard is about the space's fuel-vapor exposure, not the circuit's amperage.
  • Anode over-conservatism — assuming a bigger anode is always safer, when an oversized or wrong-alloy anode for the water type (zinc in fresh water, for instance) can create its own galvanic mismatch and accelerate corrosion of the metal it was meant to protect.
  • Diagnosing a hard-starting complaint mechanically (plugs, carb rebuild) before checking fuel age and ethanol exposure, on a boat that sat all winter with a half-full tank and no stabilizer.
  • Topping off milky gear oil instead of treating it as confirmed seal-failure evidence and pressure-testing the case to find which seal before it takes out a bearing.
  • The overcorrection: pressure- or vacuum-testing every lower unit at every routine service regardless of complaint, turning a scheduled oil change into an hour of diagnostics the job didn't call for.
  • Rushing the winterization sequence under end-of-season volume pressure and skipping a raw-water passage drain/antifreeze check that produces no symptom until a cracked block at spring commissioning.
  • Upsizing an anode in response to fast consumption without first checking hull potential and bonding continuity, masking a stray-current fault that keeps damaging through-hulls and wiring even as the bigger anode "holds."

Worked example

Situation. 32-ft express cruiser, single Volvo Penta DuoProp sterndrive (aluminum outdrive housing), kept in a brackish-water marina slip. In-water time this season: 150 days ≈ 3,600 hours. Marina labor rate: $145/hr. Customer complaint: the OEM two-anode kit (1.4 lb zinc total) is about 75% consumed after one season, and he wants "the biggest anode kit that fits" for next year.

Naive read a generalist would produce: "The anodes are clearly working hard and getting eaten — go up a size for next season." Upsized kit priced at $172 parts + 0.5 hr labor ($72.50) = $244.50, done same visit, no further testing.

Expert reasoning that overturns it. First, size the expected load: measured wetted aluminum surface on the outdrive (housing, torpedo, skeg, trim tabs) ≈ 6.5 sq ft. Using 2.5 mA/sq ft (mid-upper end of ABYC E-2's 0.2–3.0 mA/sq ft aluminum range, justified by the unpainted housing and brackish exposure): required current = 6.5 × 2.5 mA = 16.25 mA = 0.01625 A. Over the season's 3,600 immersion hours: 0.01625 A × 3,600 hr = 58.5 Ah consumed at the design rate. Zinc usable capacity at ~90% utilization efficiency: 368 Ah/lb × 0.9 = 331.2 Ah/lb. Expected anode weight consumed at design load: 58.5 Ah ÷ 331.2 Ah/lb ≈ 0.177 lb — about 13% of the 1.4 lb kit.

Actual consumption reported: 75% of 1.4 lb = 1.05 lb — roughly 5.9x the calculated baseline (1.05 ÷ 0.177 ≈ 5.9). That gap is the finding: this anode isn't undersized, it's being drawn down by something beyond passive galvanic protection.

Hull-potential reading vs. a silver/silver-chloride reference electrode: −0.95V, within the adequately-protected band (≈−0.80 to −1.10V for aluminum) — confirming protection level is fine, which rules out "anode too small to protect the metal" even though it's being consumed fast. Isolating circuits one at a time during the survey, the potential shifted measurably more negative the instant the bilge-pump breaker was switched on, identifying the pump's float-switch wiring — chafed against an adjacent through-hull fitting — as a stray-DC leakage path energizing whenever the automatic float circuit was live.

Why the naive fix is worse, not just incomplete. Doubling the anode would not stop the leakage current; it would only extend how long the anode "keeps up" with a corrosion-driving fault it isn't the cure for. Left unaddressed, the same stray current that's eating the anode 5.9x faster than expected is also driving electrolytic attack on the through-hull fitting itself — a failure mode with no anode-size ceiling and a haul-out or hull-integrity consequence the anode swap does nothing to prevent.

Actual repair and invoice:

  • Diagnostic (hull-potential survey with circuit isolation): 1.0 hr × $145 = $145.00
  • Chafed wiring repair (isolate float-switch circuit, replace splice, reroute with loom clear of the through-hull, retest continuity): 0.75 hr × $145 = $108.75
  • OEM anode kit replacement (routine, since 75% consumed — replace before next season regardless): parts $86.00 + 0.5 hr × $145 = $72.50
  • Total: $412.25

The service note, as written (deliverable, quoted):

RO — 32' Express Cruiser, Volvo Penta DuoProp sterndrive. Complaint: Anode kit ~75% consumed (1.05 lb of 1.4 lb kit) after one 150-day season; customer requested oversized replacement kit. Finding: Calculated seasonal protection load for this drive's 6.5 sq ft wetted aluminum area is ≈0.18 lb zinc at design current density — actual consumption was ~5.9x that figure. Hull potential measured −0.95V vs. Ag/AgCl, within the adequately-protected range, ruling out anode undersizing as the cause. Circuit-isolation survey traced the excess draw to the bilge-pump float-switch wiring, chafed against a through-hull fitting, creating a stray DC leakage path whenever the float circuit was energized. Repair performed: Isolated and repaired the chafed circuit; rerouted wiring clear of the through-hull; retested bonding continuity — clean. Replaced anode kit with OEM two-anode set (due for replacement regardless of the fault, given consumption level). Declined: Oversized/non-standard anode kit — would not address the stray-current source and risks accelerated corrosion of adjacent through-hull hardware from continued leakage current. Labor: 1.0 hr diagnostic + 0.75 hr wiring repair + 0.5 hr anode service = 2.25 hr @ $145/hr = $326.25. Parts: $86.00 (anode kit). Total: $412.25. Verification: Bonding continuity re-checked clean with bilge pump circuit energized; anode replacement documented against next-season baseline for comparison.

Going deeper

  • references/playbook.md — load for the filled anode-sizing worksheet, gearcase pressure-test specs by brand, compression-spec ranges by engine class, and the winterization/commissioning sequence checklist.
  • references/red-flags.md — load when reviewing a service complaint, a corrosion/anode pattern, or a pre-delivery inspection for signs of an unconfirmed diagnosis or a compliance gap.
  • references/vocabulary.md — load when a term (bonding system, phase separation, ignition protection, stray current) needs a precise definition and the misuse a generalist would make.

Sources

  • ABYC (American Boat and Yacht Council) E-2, Cathodic Protection — anode weight and current-density formulas (1.5–5.0 mA/sq ft painted steel, 0.2–3.0 mA/sq ft painted aluminum, ~368 Ah/lb zinc energy content) cited in the worked example.
  • ABYC E-11, AC and DC Electrical Systems on Boats, and SAE J1171 (External Ignition Protection of Marine Electrical Devices) — the ignition-protection requirement for any electrical component in a gasoline engine or fuel-tank space; enforced federally via 33 CFR 183.410(a).
  • ABYC H-24/H-25/H-33 — gasoline and fuel-system installation standards referenced for fuel-space ignition-protection scope.
  • Mercury Marine technical guidance on E10 ethanol-blend fuel — the roughly 0.5% water-by-volume phase-separation threshold at typical ambient temperature (lower at cold temperatures) and the guidance that stabilizer treatment cannot reverse separation once it occurs.
  • OEM service manuals (Yamaha, Mercury, Volvo Penta, Mercruiser) — drive-specific gearcase pressure-test hold specs (commonly 10–15 psi, with some OEM specs at 14 psi held 2 minutes) and vacuum-test procedures (5–10 inHg) for lip-seal-specific checks, cited as representative of published ranges rather than one live lookup.
  • Practitioner forums (The Hull Truth, iboats) — common trade discussion and consensus on pressure-test technique and milky-oil-as-seal-failure diagnosis; cited as practitioner consensus, not a single authoritative document.
  • No direct motorboat-mechanic practitioner has reviewed this file yet — flag corrections or gaps via PR.

What ships with it: 3 files

23.0 KB alongside SKILL.md

references/

Keep looking

Skills are one crate of 613,757. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.