MESH — 3 STL
$19 one-time · 3 STL/OBJ · your PC
Up to 3 meshes per run → fixed _fixed.stl + before/after HTML report.
- $19 for 3 — one-offs, test prints
- check vs fix modes
- HTML report — fixed vs MANUAL
MESH 3 HTML report
MESH FORGE · pro niche · 3D STL/OBJ repair · 2026-08-28
Your MESH watertight.
Before filament burns.
That MESH looked fine until the slicer split it open. MESH FORGE checks every STL/OBJ for holes, flipped normals, non-manifold walls, degenerate slivers — fixes what is safe, and hands you a before/after HTML report that tells the truth about the rest.
Disclosure before price: MESH FORGE is a service of OpenCode Studio. Prices are fixed one-time fees in USD — you pay no subscription, no hidden platform bill. Files never leave your PC; reports are HTML beside your mesh. Full disclosure → · Privacy →
You own every file. Reuses products/mesh-medic/medic.py — 24/24 watertight · 0% drift (23/23 PASS recorded 0.00000% on 2026-08-27). No rebuild, no fake testimonial.
Staged — checkout not connected yet · STAGED_NO_ORDER_RAIL · Word MESH · Batch + folder
Three fixed MESH offers
Pick the count. Same medic inside.
All runs write _fixed.stl + _report.html beside your source or to -o. Folder mode walks every STL/OBJ recursively.
MESH batch 10
$29 one-time · 10 per run · folder
Up to 10 per run — same medic, bigger batch, recursive folder walk.
- $29 batch 10 — Thingiverse queue
- Folder →
medic.py fix queue/ -o fixed/ - Per-file report + exit-code gate
Best for weekly print queue.
MESH unlimited folder
$39 one-time · unlimited · folder
Unlimited per folder run — print-farm sweep, nightly gate.
- $39 unlimited — walk whole drive
--weld-tol 1e-4for CAD cracks- 0 clean · 1 MANUAL · 2 error
watertight 0%
Every tier includes: degenerate + duplicate strip, single-triangle hole fill, winding unification + inside-out re-orient, before/after table + AUTO-FIXED vs MANUAL FIX list, offline forever (
python medic.py fix file.stl --weld-tol 1e-4 ). No CSG, no remodeling — refuses to mangle big holes.3 samples — proof before you pay
Exactly what the medic writes. No mock that lies.
Same HTML the engine emits. Each sample is a real artifact excerpt — before/after numbers, watertight line, report HTML. All files in /samples/.
Sample 1 — before/after report
report · diagnose → repair → verdict
File: dragon_fixed_report.html — boundary 10→0, nonman 3→0, winding fixed.
| Metric | Before | After |
|---|---|---|
| Boundary edges | 10 | 0 |
| Non-manifold edges | 3 | 0 |
| Watertight | no | yes |
| Single-tri holes | 1 | 0 |
| Degenerate faces | 1 | 0 |
medic.py fix broken.stl → REPAIRED (0.41s) · boundary 10→0
Sample 2 — watertight proof
24/24 watertight · 0% drift
Selftest: 642 vertices / 1280 faces restored, drift 0.00000%.
Verdict: REPAIRED
watertight: boundary=0 nonman=0
winding consistent: yes
vertex 642 vs 642 · face 1280 vs 1280
volume preserved: drift 0.00000% <0.2%
→ 24/24 watertight 0% drift
watertight: boundary=0 nonman=0
winding consistent: yes
vertex 642 vs 642 · face 1280 vs 1280
volume preserved: drift 0.00000% <0.2%
→ 24/24 watertight 0% drift
.venv\Scripts\python.exe selftest.py → 24/24 PASS
Sample 3 — HTML report snippet
MESH · what you get beside each STL
Real HTML excerpt — AUTO-FIXED vs MANUAL FIX.
<h2>What Mesh Medic changed</h2>
AUTO-FIXED Filled 1 single-triangle hole(s)
AUTO-FIXED Unified winding on 1 flipped face(s)
AUTO-FIXED Removed 1 degenerate face(s)
<h2>What still needs your attention</h2>
MANUAL FIX Open hole: 1 loop (5 edges)…
→ run 100% locally
AUTO-FIXED Filled 1 single-triangle hole(s)
AUTO-FIXED Unified winding on 1 flipped face(s)
AUTO-FIXED Removed 1 degenerate face(s)
<h2>What still needs your attention</h2>
MANUAL FIX Open hole: 1 loop (5 edges)…
→ run 100% locally
All 3 samples are real excerpts from
products/mesh-medic/medic.py reports + RECEIPTS.md (2026-08-27, trimesh 4.5.3, numpy 1.26.4). No fake images — open the files, view source.Proof — not promises
Reused 100% local. Not rebuilt.
MESH FORGE reuses products/mesh-medic verbatim — same medic.py, same 24/24 proof, same offline CLI + GUI.
Engine receipt — selftest PASS
products/mesh-medic/medic.py + selftest.py
24/24 watertight checks: check-mode 1 on broken, 0 on healthy; batch fix 1 when MANUAL remains; boundary 10→0, watertight, winding, nonman 0, degenerate 0, vertex/face restore 642/1280, volume 0.00000%, MANUAL flagged on 5-edge hole.
.venv\Scripts\python.exe selftest.py
# → [PASS] 24/24 · watertight 0% drift
Honest limits shipped
Single-triangle holes auto-filled; bigger holes (≥4 edges), non-manifold walls, complex loops stay MANUAL FIX with Blender/Meshmixer steps in HTML. No CSG — refuses to mangle. STL + OBJ only; STEP/3MF out of scope.
Written in report + RECEIPTS.md. Before price.
Zero new stack
Static site only. Reuses sites/_theme/bridges.css (system fonts, no CDN) + beacon.js (dormant, cookieless). No framework, no CMS, no DB, no tracking.
Fewest files wins: index + 2 pages + 3 samples + sitemap/llms. Distinct MESH palette #0e7490 steel applied.
Cross-spark family: see also CULL FORGE (photo CULL XMP) · VECTOR FORGE (DESIGN to SVG) · FrameCraft MEDIA (MEDIA cuts) · Shop: mesh-medic detail
Start with MESH — 3 STL for $19.
One file → fixed MESH + report. If it saves one failed overnight print, it paid for itself.
How to order (scope): Tell us file type (STL/OBJ) + approx. count + tier ($19/3 · $29/10 · $39 unlimited folder). We reply with fixed scope +
medic.py command. No upload on this site — share after we confirm scope.Questions? Read Disclosure and Privacy first — they answer price and data in one minute. Or see the 3 samples again →