arch-lookup

Cross-reference Tenstorrent silicon specifications with tt-emule mock-API implementations.

4|1|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/tenstorrent/tt-emule --skill arch-lookup
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: arch-lookup
Source: https://github.com/tenstorrent/tt-emule/tree/main/.claude/skills/arch-lookup
Command: npx skills add https://github.com/tenstorrent/tt-emule --skill arch-lookup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill resolves the ambiguity between complex Tenstorrent hardware specifications and their software-emulated counterparts, ensuring mock implementations remain faithful to silicon behavior.

Core Features & Use Cases

  • Multi-Arch Orchestration: Simultaneously queries architecture-specific agents (Wormhole, Blackhole, Quasar) to gather authoritative hardware data.
  • Emule Mapping: Automatically cross-references hardware findings with existing tt-emule mock implementations to identify gaps or divergence.
  • Use Case: When implementing a new kernel op, use this skill to verify if the hardware supports specific data formats and receive a recommended strategy for stubbing the API in the emule runtime.

Quick Start

Use the arch-lookup skill to determine how the unpack engine handles Float16 on Blackhole and identify the necessary mock implementation strategy.

Frequently Asked Questions about arch-lookup

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

FAQPage Schema
How do I verify if my tt-emule mock-API implementation matches Tenstorrent hardware specifications?

To verify mock-API implementations against hardware specifications, this skill cross-references authoritative silicon data from architecture-specific agents with existing tt-emule mock implementations to identify architectural divergences and provide implementation guidance.

How does architecture-specific hardware emulation handle Wormhole and Blackhole divergence?

Architecture-specific hardware emulation handles divergence by orchestrating parallel agent queries for Wormhole, Blackhole, and Quasar architectures, gathering authoritative hardware data to synthesize actionable mock-API recommendations for the emule runtime.

What is the best way to implement software-emulated kernel operations for Tenstorrent architectures?

The best way to implement software-emulated kernel operations is by querying architecture-specific agents to retrieve authoritative silicon specifications and cross-referencing them with tt-emule mock-API implementations to ensure faithful silicon behavior.

Does the arch-lookup skill require access to internal documentation and the tt-emule codebase?

Yes, the arch-lookup skill requires integration with internal documentation sources and the tt-emule codebase to synthesize actionable mock-API recommendations and identify gaps between hardware specifications and software-emulated counterparts.

Why does my emule mock-API implementation diverge from actual silicon behavior?

Emule mock-API implementations diverge from actual silicon behavior due to ambiguity between complex Tenstorrent hardware specifications and their software-emulated counterparts, which this skill resolves by identifying architectural gaps and providing implementation guidance.