trtllm-flashinfer-upgrade

Automate flashinfer-python version upgrades across TensorRT-LLM pinned files.

Updated May 23, 2026
One-click install
npx skills add https://github.com/yo-steven/skills-exploration-20260522 --skill trtllm-flashinfer-upgrade
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trtllm-flashinfer-upgrade
Source: https://github.com/yo-steven/skills-exploration-20260522/tree/main/skills/TensorRT-LLM/trtllm-flashinfer-upgrade
Command: npx skills add https://github.com/yo-steven/skills-exploration-20260522 --skill trtllm-flashinfer-upgrade

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Upgrading flashinfer-python in TensorRT-LLM is error-prone because versions and metadata are pinned across multiple files and may require careful lockfile handling.

Core Features & Use Cases

  • Release-aware upgrade planning: Fetches stable and optionally nightly flashinfer releases from GitHub and compares them to the current pinned version in requirements.txt.
  • Repo-wide version reference updates: Updates all relevant version pins, including requirements.txt, security_scanning/pyproject.toml, and ATTRIBUTIONS-Python.md.
  • Optional lockfile synchronization: Can update security_scanning/poetry.lock hashes and regenerate poetry metadata when the user explicitly opts in.
  • Validation and PR workflow: Checks for version-gated compatibility logic and guides the commit, push to a fork, and PR creation steps.

Quick Start

Tell the skill to upgrade flashinfer-python to the latest stable release and skip updating security_scanning/poetry.lock for minimal diffs.

Frequently Asked Questions about trtllm-flashinfer-upgrade

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

FAQPage Schema
How do I upgrade flashinfer-python in TensorRT-LLM safely?

To upgrade flashinfer-python in TensorRT-LLM safely, fetch available GitHub releases, select a target version, and update all pinned references across requirements.txt, security_scanning/pyproject.toml, and ATTRIBUTIONS-Python.md.

What files need to be updated when bumping the flashinfer-python dependency?

When bumping the flashinfer-python dependency, you must update version pins in requirements.txt, security_scanning/pyproject.toml, and ATTRIBUTIONS-Python.md to keep scanning metadata and attributions consistent.

Can I regenerate poetry.lock hashes during a flashinfer-python upgrade?

Yes, you can regenerate security_scanning/poetry.lock hashes during a flashinfer-python upgrade by explicitly opting in to lockfile synchronization to update poetry metadata safely.

Do I need GitHub CLI authentication to upgrade flashinfer-python versions?

Yes, GitHub CLI authentication is required to fetch available flashinfer releases and guide the commit, fork push, and PR creation workflow for the version upgrade.

What is the best way to handle routine dependency maintenance for flashinfer-python?

The best way to handle routine flashinfer-python dependency maintenance is automating release-aware upgrade planning that compares stable and nightly GitHub releases to current pinned versions before applying repo-wide updates.

Why does upgrading flashinfer-python in TensorRT-LLM fail without coordinated edits?

Upgrading flashinfer-python fails without coordinated edits because versions and metadata are pinned across multiple files, requiring consistent updates to requirements, security scanning configurations, and attributions.