estampo/estampo v0.3.0

on GitHub

Features

• Add CuraEngine slicer backend (engine = "cura") as alternative to OrcaSlicer (#282)
• Build CuraEngine from source instead of extracting from AppImage, eliminating placeholder G-code header issue (#299)
• Add Bambu Lab P1S machine definition for CuraEngine with proper start/end G-code (#301)
• Add OrcaSlicer 2.3.2 Docker image builds alongside 2.3.1
• Add --engine cura support to profile extraction script and CI workflow for automated CuraEngine printer definition discovery
• Add EngineConfig base class, SlicerConfig.active property, top-level [filaments] table for material aliases, and PartConfig.material field; remove facade fields from SlicerConfig
• Add [slicer.filament_overrides] to patch all filament profiles (e.g. required_nozzle_HRC)
• Add [slicer.machine_overrides] for overriding machine profile settings (e.g. nozzle_type = "hardened_steel" for carbon-fiber filaments).
• Add bed_type config option to set plate type for filament compatibility
• CuraEngine machine definitions are now stored as JSON files (bundled Bambu Lab P1S 0.4 nozzle.json); add custom printers via estampo profiles add; profiles pin and profiles add no longer misroute Cura profiles; override values from config are now coerced to the correct type.
• CuraEngine printer definitions: discoverable selection during estampo init, configurable via [slicer.cura] printer, and dynamic definition resolution at slice time
• Engine-namespaced TOML config: [slicer.orca] and [slicer.cura] sections for per-engine settings with backward-compatible legacy format support
• Engine-namespaced profile directories: profiles/<engine>/<category>/ with backward-compatible legacy path fallback
• Extract Bambu Connect 3MF fixup into a packaged_output pipeline stage so printer-specific packaging is explicit and only runs for Bambu printers
• Show animated status spinners during slow Docker operations (image pull, profile extraction)
• Support CuraEngine in local slicer discovery (find_slicer) for all platforms
• estampo init wizard now prompts for slicer engine (OrcaSlicer or CuraEngine) and adapts version detection and profile discovery accordingly.

Bugfixes

• Fix CuraEngine Docker image breaking Python by isolating bundled libssl/libcrypto (#283)
• Fix CuraEngine Docker image Python access for unprivileged estampo user (#286)
• Show CuraEngine error output instead of suppressing stderr (#288)
• Show rich UI output (checkmarks, print summary) alongside debug logs in verbose mode (#300)
• Fix ValueError when CuraEngine override values include a % suffix (e.g. sparse_infill_density = "35%"). (#310)
• Fix profiles pin failing for CuraEngine when printer name includes nozzle suffix or different casing (#340)
• Fix CuraEngine slice failing when no separate machine profile JSON exists for the printer (#345)
• Fix CuraEngine squashed definitions preserving inherits for unresolved base definitions like fdmprinter (#347)
• Fix profile extraction push failure when bot-managed branch already exists (#363)
• Remove Bambu-specific AMS auto-detect from estampo init wizard (#367)
• Skip printer setup and print stage for CuraEngine in estampo init (#380)
• Broadcast scalar machine overrides to arrays when profile has per-extruder fields
• Error when pinned profiles don't match the target slicer version instead of silently crashing
• Fix CuraEngine binary "not found" by patching ELF interpreter and replace missing bambu_3mf dependency with inline G-code generation
• Fix CuraEngine reporting zero filament usage — CuraEngine CLI writes placeholder G-code header values that are not updated after slicing; patch with real values from stderr and fall back to E-value analysis.
• Fix CuraEngine slicing by converting 3MF model data to STL before passing to the engine
• Fix CuraEngine slicing: place mesh on bed, resolve G-code template variables, fix header patching
• Fix CuraEngine: disable default brim, center mesh on build plate
• Fix OrcaSlicer 2.3.2 segfault in Docker on Colima by preventing Wayland GL init
• Fix TestPyPI dev version collisions by using git commit count instead of workflow run number.
• Fix filament weight reporting: correct multi-value regex, M83 relative extrusion support, newest-file selection, and layer count in output summary.
• Fix profile extraction push failure when branch already exists from earlier run
• Fix release workflow checkout ref that prevented manual release dispatch
• Fix release workflow job skip cascade on manual dispatch
• Fix release workflow tag-check that used curl flags with gh api
• Narrow bare except Exception handlers to specific exception types for better debuggability.
• Pass --allow-mix-temp for multi-filament AMS configurations on OrcaSlicer 2.3.2
• Pass --allow-mix-temp unconditionally on OrcaSlicer 2.3.2+ to fix grouping errors; gate the flag off for older versions.
• Replace unsafe eval() in CuraEngine G-code template substitution with AST-based safe evaluators.
• Revert xvfb-run wrapping that caused OrcaSlicer 2.3.2 to segfault on GL init.
• Set infill_line_distance directly so CuraEngine respects infill density overrides
• Stage Docker slicer input inside output directory and preserve file extension to fix bind-mount issues

Misc

• Run e2e slice test from host to match real user flow (Docker-based slicing). (#275)
• Fix lint errors in scripts/bambu_cloud_login.py and scripts/test_cloud_print.py (#331)
• Populate bundled CuraEngine 5.12.0 printer definition manifest with full machine list extracted from Docker image (#354)
• Add ADR-006 defining the slicer plugin protocol: each engine is a self-contained module; top-level files are pure dispatch layers; clear contract for adding new slicers
• Add Architecture Decision Records, ROADMAP.md, and module ownership table to CLAUDE.md for long-term architectural coherence
• Add GitHub issue templates, labels, and milestones for structured backlog tracking
• Add scripts/compare_engines.py utility to compare OrcaSlicer vs CuraEngine G-code output
• Add automatic Claude PR review workflow — checks ADR compliance, exception handling, eval() usage, and missing changelog fragments on every PR
• Add slicer engine research doc: Bambu P1S multi-material support across CuraEngine, PrusaSlicer, and Kiri:Moto.
• Consolidate bambu-3mf + bambu-cloud into single bambox package in docs and ADR-005
• Disable OrcaSlicer 2.3.2 builds (segfaults on slice); re-enable when upstream fixes it
• Extract OrcaSlicer-specific logic from slicer.py, profiles.py, and init.py into new orca.py engine module per ADR-006.
• Extract shared magic numbers into constants.py and migrate example configs to engine-namespaced [slicer.orca]/[slicer.cura] format
• Fix Claude PR review workflow: add missing id-token: write permission for OIDC authentication
• Fix Claude PR review workflow: raise max-turns to 8 and add continue-on-error so turn limit never blocks a merge
• Fix release-readiness jobs skipping on manual dispatch and nightly schedule
• Improve CuraEngine error reporting to include both stdout and stderr
• Mark all todo-slicer.md items as resolved — slicer-agnostic work is complete.
• Move CuraEngine definition pinning and profile logic from profiles.py into cura.py engine module.
• Profile system now handles CuraEngine gracefully — profiles list --engine cura explains that CuraEngine uses inline settings instead of showing empty results.
• Remove Claude PR review workflow — not working reliably
• Remove legacy flat [slicer] config format, fabprint.toml alias, FABPRINT_* env vars, and ~/.config/fabprint migration.
• Replace hardcoded OrcaSlicer references with engine-agnostic text in CLI help, error messages, and 3MF metadata
• Speed up CI with uv dependency caching, concurrency groups, and conditional bridge builds
• Update docs and README to use engine-namespaced [slicer.orca]/[slicer.cura] config format and document [filaments] table
• Use RELEASE_PAT in prepare-release workflow so CI triggers on release PRs automatically