# Pi Pods

> Pi Pods

- Skill: `romiluz13/pi-pods` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add romiluz13/pi-pods`
- Raw SKILL.md: https://api.skillmd.com/api/skills/romiluz13/pi-pods/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: romiluz13 (https://skillmd.com/u/romiluz13)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/romiluz13/pi-pods

---


# Pi Pods

## Grounding

1. `pi-mono/packages/pods/README.md` — installation, pod management, model commands, GPU multi-assignment, and pre-defined models.
2. `pi-mono/packages/pods/src/` — CLI commands implementation if needing deeper args validation.

## Invariants

- **Auto-assignment**: When running multiple models on the same pod, `pi` automatically assigns them to different GPUs.
- **Parameter Ignorance**: When passing custom vLLM args with `--vllm`, the default CLI shortcuts for `--memory`, `--context`, and `--gpus` are ignored.

## Workflows

- **Setup Pod**: Use `pi pods setup <name> "<ssh>"` along with `--mount` for shared NFS storage (DataCrunch) or network volumes (RunPod).
- **Start Pre-defined Model**: Use `pi start <model> --name <name>` for known agentic models (Qwen, GLM, GPT-OSS). The tool calling parsers are automatically configured.
- **Custom vLLM Args**: Pass specific settings (e.g. tensor parallelism) using `--vllm --tensor-parallel-size <N>`.

## Anti-patterns

- Do not manually construct tool-calling parsers for pre-defined models like Qwen or GLM; `pi` configures `hermes` or `glm4_moe` automatically.
- Do not assume models are downloaded redundantly on DataCrunch; emphasize the NFS shared models path (`/mnt/hf-models`).