mirror of https://github.com/Heretek-AI/CachyLLama-BUILDER.git synced 2026-10-11 20:59:54 +00:00

Files

T

John Smith 944f56ba73 Remove Windows and macOS builds from CI/CD workflows

- Remove build-bundle-macos from build-llama-ai-bundle.yml.
- Remove windows-cpu and macos-metal from build-cachyllama-all.yml.
- Remove build-windows from build-cachyllama-rocm.yml.
- Remove test-windows from test-cachyllama-rocm.yml.
- Streamline all workflows, scripts, and documentation for Linux builds.

2026-08-30 20:49:55 -04:00

1.4 KiB

Raw Permalink Blame History

GEMINI.md

Developer reference for Google Gemini / Antigravity agents working in CachyLLama-BUILDER.

🤖 Antigravity & Gemini Operational Protocols

Workspace Context

Repository: Heretek-AI/CachyLLama-BUILDER
Upstream Targets:
- fewtarius/CachyLLama (C++ inference engine fork)
- fewtarius/llama-ai (APU/GPU orchestrator and solver)
Reference Architectures:
- lemonade-sdk/llamacpp-rocm
- lemonade-sdk/llama.cpp

Core Development Principles

Never Break CI Portability: Standalone binaries must function without system-installed ROCm or CUDA drivers by bundling all required shared objects and setting RPATH=$ORIGIN.
AMD APU Priority: Strix Halo (gfx1151), Strix Point (gfx1150), and Phoenix (gfx110X) are primary targets. Always test that shader compile flags, subgroup configurations, and memory allocators align with CachyLLama defaults.
Planning Protocol: For complex workflow restructuring, always use the /plan workflow: create a detailed implementation plan artifact, seek user review, verify with automated checks, and generate a walkthrough.

Quick Verification Suite

# Validate YAML workflow schemas
python3 -c "import yaml, glob; [yaml.safe_load(open(f)) for f in glob.glob('.github/workflows/*.yml') + glob.glob('.github/actions/**/*.yml', recursive=True)]"

# Check Python and shell scripts
python3 -m py_compile scripts/*.py && bash -n scripts/*.sh