LouPineWays

context-clearing

Verify that a Loop-Dee-Loup PR, issue, or parent snapshot is a self-sufficient durable handoff before clearing, compacting, yielding, or replacing a session.

local-worker

Invoke a configured local LLM for a bounded, cheaply-verifiable subtask a Claude subagent has already decided is a good local-delegation candidate.

model-check

Recommend the cheapest safe model, effort, and execution shape for a Loop-Dee-Loup slice before dispatch. Advisory only.

spend

Diagnose where a Loop-Dee-Loup session spent tokens and whether that spend produced validated progress.

Diagnose and fix a reported false-positive or false-negative in Stage 1's genuine-response classifier (tools/review-watch/genuine-response.mjs) — isGenuineResponse, findExemption, and the primitives they compose.

telemetry-battery

Run the approximately-weekly /spend + maker telemetry validation battery, gating optimization conclusions on telemetry coverage rather than assuming instrumentation is sufficient.

persona-maker

Create a compact, reusable expertise profile for a recurring class of subagent work, stored under `.claude/personas/`.

script-maker

Convert a demonstrated, repeated deterministic repository operation into a script, replacing model reasoning with a checked, verifiable tool.

skill-maker

Create a new repository-local skill when a recurring reasoning or workflow problem has become stable enough to encode, and no existing mechanism already solves it.