Talk to your logistics emissions data in plain language.

This repository contains the iLEAP Skill, which lets you interact with the iLEAP API using tools such as Claude.ai, Claude, or similar software offerings.

Status: community preview. The skill works today, and we are sharing it early to gather feedback before committing to a stable release. Expect changes between now and then.

You can download the latest version of the iLEAP Skill here.

• Software engineers: a Rust CLI plus a packaged skill you can run, inspect, and use as a template (see Technical Details).
• Product managers: a community preview that shows what natural-language access to logistics emissions data looks like (see What to expect).
• Sustainability managers: a no-lock-in way to explore iLEAP-aligned emissions data and understand how it is calculated (see the iLEAP Initiative).

The iLEAP skill lets you explore transport emissions data (shipments, Transport Operation Categories (TOCs), Hub Operation Categories (HOCs), footprints, and more) just by asking Claude in plain language. No spreadsheets, no API calls. Ask for a dashboard, a what-if scenario, or an export, and Claude fetches the data and builds it for you.

If you are looking for some example outputs, you can find them on the skill's website at https://skill-preview.ileap.dev.

The iLEAP Initiative exists to decarbonize logistics at scale, using digitalization to get there. The iLEAP Skill is one way we put that mission into practice.

The skill builds on a core deliverable of the initiative, the iLEAP Technical Specifications. These describe how to make logistics emissions data available, how the data is structured, and what it means. That last part matters most: it covers how the data must be calculated to be representative and complete, how its quality can be assessed, and more.

The Specifications are an interoperability standard, and that brings a concrete benefit: there is no lock-in. Any tool provider can implement them, and anyone can use this skill to work with iLEAP-aligned data right away, no matter who produced it.

The following instructions apply to Claude.ai only, the only platform we have tested so far.

You can find them right on the iLEAP Skill's website at https://skill-preview.ileap.dev.

Where we are. The iLEAP Skill is under active development and offered as a community preview. Please do not use it for production purposes yet.

What is safe to do. Explore iLEAP data, try out queries, and share feedback. Because AI can make mistakes, always review the skill's output before you rely on it. When you pull in data, make sure you have the rights to process it with the skill.

Good to know.

• This is not an endorsement of Claude, Anthropic, or any other AI approach or vendor.
• The skill has only been built for and tested with Claude/Anthropic. It should work with other providers, including local AI deployments, too.
• AI use carries a significant environmental impact. We take this seriously and weigh it against the emissions reductions the iLEAP Initiative is working to unlock.

We would love to hear from you. Please send us your feedback, and check back for updates.

1. The iLEAP CLI is implemented in Rust (see src/).
2. The skill uses prebuilt static Linux binaries of it. With every release, an ileap-skill.zip with pre-built binaries is created.
3. To build them locally and after cloning the repo, you can either run cargo build --release, or cargo install --path . to install the CLI on your machine.
4. The skill is available in .agents/skills/ileap and can be used as a template for building your own skills on top of the iLEAP CLI.