Sync all three vLLM model configs from club-3090 master (ae4846f).
Update to Genesis v7.65 full PROD env set with new patches.
Update docker image to nightly-7a1eb8ac. Add torch_compile and
triton cache dirs. Add agent setup guide (AGENTS.md).
Add 'evan' API key to llama-swap sops secrets.
Allow one CUDA0 and one CUDA1 model to run simultaneously. Dual-GPU
models (using -ts splits) are excluded from the matrix so they evict
everything when loaded. vLLM docker models get evict_cost=50 to
discourage eviction due to slow cold starts.
Extract a shared hasType helper for model filtering and add
vision (text + image) input capability to compatible models.
Also tag two llama-swap models with the vision type.
- Add modules/home/security/pass-keyring with GPG agent, pass, and
python keyring backend config for headless credential storage
- Enable pass-keyring for lin-va-mbp-work-vm
- Update bash PATH from ~/.bin to ~/.local/bin
Allow LSP servers to be enabled on-demand via a buffer-local command
instead of auto-starting on matching filetypes. The command name is
auto-derived from the server name (e.g. 'GolangciLint'). Switch
golangci-lint to manual mode as it's resource-heavy and not always needed.
Add per-item blur_radius=30 with background.clip=0.5 to create a
frosted-glass effect on each component pill while keeping the bar
itself fully transparent. Bump bg alpha from 0x20 to 0x30 so the
blur is visible through the semi-transparent backgrounds.
- Add new address-gh-review skill with SKILL.md and gh_review.sh script
for fetching and addressing unresolved GitHub PR review comments
- Update AGENTS.md comment style to require Title Case for block titles
- Add packages/nunc derivation built from Gitea with swiftPackages
- Add home module reichard.services.nunc with configurable settings
and launchd agent for macOS
- Enable nunc service on mac-va-mbp-work
- Tweak mac-va-mbp-work defaults: hide menu bar, disable auto
brightness, hide Finder desktop icons
Add a new shell function that uses an LLM API to generate shell commands
from natural language queries. The command displays the generated command
for user confirmation before execution and adds it to history.
- Generates shell commands from natural language via LLM
- Displays command for user review before execution
- Automatically adds executed commands to bash history
- Uses qwen3-coder-next-80b-instruct model
