skills.rest
diazMelgarejo
diazMelgarejoCOMMUNITY
Explore all skills by this author →
Website𝕏@Melgarejo_yDiazGitHub

local-inference

Routes local inference decisions to the canonical hardware-affinity-gate skill card.

Updated Mar 22, 2026
Search Tags:#git sync#skill routing#local inference#hardware affinity#orama-system#canonical skill
One-click install
npx skills add https://github.com/diazMelgarejo/orama-system --skill local-inference-diazmelgarejo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: local-inference
Source: https://github.com/diazMelgarejo/orama-system/tree/main/.agents/skills/local-inference
Command: npx skills add https://github.com/diazMelgarejo/orama-system --skill local-inference-diazmelgarejo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents stale or duplicated local-inference routing logic by acting as a thin wrapper that always points to the canonical hardware-affinity-gate skill, ensuring agents follow a single source of truth for hardware policy. ## Core Features & Use Cases - Canonical Redirection: Loads the authoritative hardware-affinity-gate SKILL.md from the repo instead of duplicating behavior. - Freshness Checks: Runs git fetch, status, and conditional fast-forward pull before reading the canonical card, reporting drift when the worktree is dirty. - Use Case: An agent asked to decide whether a model should run locally loads this wrapper, syncs the repo, and follows the canonical hardware-affinity-gate card plus the pt-hardware-policy command edge. ## Quick Start Ask the agent to load the local-inference skill and follow the canonical hardware-affinity-gate card before making any local inference routing decision.

Frequently Asked Questions about local-inference

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I route local inference decisions in the orama-system repo?▼

Load the local-inference wrapper, which directs you to the canonical hardware-affinity-gate SKILL.md under bin/orama-system/skills. Follow that card and the pt-hardware-policy command edge rather than copying behavior from the wrapper.

What is the hardware-affinity-gate skill?▼

It is the canonical skill that absorbed local-inference routing logic within the orama methodology. The local-inference directory is only a thin wrapper pointing to it, with PT runtime policy handled by the pt-hardware-policy command.

Does local-inference work without a git repository?▼

The wrapper resolves the repo root via git rev-parse, falling back to the current directory. Without git tracking it cannot verify freshness, so it reports drift and reads the canonical card with that caveat.

Why does the wrapper run git fetch before loading the skill?▼

It ensures the canonical hardware-affinity-gate card is current before use. If the branch tracks a remote and the worktree is clean, it fast-forward pulls; otherwise it reports drift instead of modifying the repo.

When should I not rely on the local-inference wrapper?▼

Do not copy behavior from the wrapper itself, since it contains no routing logic. If the canonical card path is missing or the repo state is uncertain, treat the drift report as a blocker and resolve it first.

Related Skills

Explore More Software Engineering Skills→
HugopeckHugopeck

update-skills

Refresh installed Track skill clones with fast-forward-only git pulls.

Community
Intermediate
IrisGoLabIrisGoLab
6

win-ai-local

Manage Ollama models and run local LLM inference on Windows 10/11.

Official
Intermediate
ShalomObongoShalomObongo
3

debug

Route legacy debug skill references to the canonical implementation.

Community
Basic
bailynlovebailynlove
1

llama-cpp

Run llama.cpp models locally and discover GGUF models from Hugging Face Hub.

Community
Intermediate