TASK-DESCRIPTION

• Create: documents/11-minerva-travel-plan.md

Plan the minerva offline laptop deployment.

Connect to minerva.lan over SSH (per the homelab inventory) and discover the hardware: GPU (lspci | grep -iE 'vga|3d', vulkaninfo --summary if Vulkan is present), RAM (free -h), and disk (df -h ~).

Create documents/11-minerva-travel-plan.md containing:

• The verified hardware facts with the date.
• The chosen travel model from the candidates in prompts/features/05-minerva-travel-laptop.md, with its exact GGUF name, size on disk, expected context, and the rationale.
• The deployment shape: llama.cpp install (backend chosen from the GPU), a single-model or small-router preset on 127.0.0.1:2001, and the aider and llama.vscode invocations pointing at http://127.0.0.1:2001/v1.
• The offline checklist: predownload the GGUF, verify the local server answers, and confirm the tools work with the laptop’s network disabled.

Do not modify any file outside documents/11-minerva-travel-plan.md. If minerva is unreachable, create the document with the hardware section marked unverified and the model choice deferred, and report the connection failure.

TASK-OUTPUT

Provide a summary of the created document and the verified hardware facts. No commentary beyond the summary.

TASK-CONTEXT

Feature requirements: prompts/features/05-minerva-travel-laptop.md. minerva.lan is minerva.lan, WiFi-attached, with SSH per homelab/sources/inventory/hosts.yaml. The homelab Ansible inventory may be used for provisioning once specs are known.

TASK-FILES

TASK-VERIFY

• Run: make test from the repository root.
• Expected: exit status 0.

TASK-FEATURES

• prompts/features/05-minerva-travel-laptop.md

TASK-ACCEPTANCE

• MINERVA-R001
• MINERVA-R002
• MINERVA-R003

OUTPUT: the content of documents/11-minerva-travel-plan.md, as TASK-OUTPUT describes, and nothing else.