eval

name	eval
description	Persistent code-mode kernels (JS/Python) for incremental computation — omp/senpi eval analog on GrokBuild via lfg MCP. Use when multi-step code execution, data exploration, or stateful REPL beats one-shot bash.

eval (code mode)

Grok-adapted code mode inspired by oh-my-pi eval and senpi senpi-codemode.

Use the eval MCP server tools (installed by lfg setup --run):

Tool	Purpose
eval	Run one cell in a persistent kernel (js or py)
eval_reset	Wipe kernel state for a language or all

When to use eval (not bash)

Prefer eval for:

• Multi-step compute (imports → define → test → use)
• Loops, conditionals, dataframes, JSON munging
• Anything where state should survive across steps
• Fragile quoting that bash would mangle

Prefer bash only for short single pipelines (ls, git status, one-shot CLI).

Discipline (omp parity)

One eval call = one logical step.

1. Imports in one call
2. Define helpers/data in the next
3. Test
4. Use

Do not re-import / re-declare names already defined unless you reset, the kernel crashed, or you got NameError / ReferenceError.

Arguments

{
  "language": "js" | "py",
  "code": "<cell body, verbatim>",
  "title": "optional label",
  "timeout": 30,
  "reset": false,
  "cwd": "/optional/workdir"
}

• reset: true wipes that language only before the cell.
• State is per (cwd, language) for the life of the MCP process.

Languages (MVP)

Token	Runtime
js	Node VM worker (top-level await ok)
py	python3 subprocess (exec in persistent globals)

Ruby/Julia kernels exist in omp/senpi but are not in this Grok MVP.

Examples

Python — incremental

eval language=py code="import json" title=imports
eval language=py code="data = {\"n\": 3}" title=define
eval language=py code="print(data[\"n\"] * 2)" title=use

JavaScript — top-level await

eval language=js code="const n = 2" title=define
eval language=js code="n + 40" title=use

Reset after poison state

eval_reset language=py

Failures

• Kernel timeout → tool returns error; next call starts a fresh kernel for that key.
• Syntax / runtime errors → ok: false with traceback; prior names still alive.
• Missing python3 → py cells fail until available (override with LFG_EVAL_PYTHON).

Out of scope (honesty)

• No tool.* / agent() / completion() bridges (omp full eval has those).
• No rich display / image mime bundles.
• No rb/jl kernels.
• State does not survive Grok process restart (in-process MCP lifetime only).