Skip to main content
skill CX Codex

jupiter-vrfd — Claude Code skill: Jupiter token verification

jupiter-vrfd — a skill for Claude Code that guides agents through Jupiter token verification on Solana. Handles eligibility checks, payment crafting, and metadata submissions costing 1000 JUP.

MIT 271 installs ★ 90 updated 2026-07-07

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add jup-ag/agent-skills --skill jupiter-vrfd -a claude-code
Open on GitHub

What jupiter-vrfd does

This skill runs the public Jupiter express token-verification pipeline on Solana. It queries https://api.jup.ag to check canVerify and canMetadata eligibility flags, crafts the 1000 JUP submission payment transaction, and submits token metadata updates.

Operational rules restrict sensitive data: the agent never prompts for private keys or seed phrases in chat, relying only on derived addresses and local signer files. If network access or local signing tooling is unavailable, the agent halts before execution and prints exact manual commands. Calls require an x-api-key from developers.jup.ag.

What it helps with

  • check Solana token eligibility for Jupiter verification
  • craft the 1000 JUP verification payment transaction
  • submit token metadata updates via express verification
  • execute the verified token submission flow

Example prompts for the agent

  • Check eligibility for verifying token JUPyiwrYJFskUPiHa7hkeR8VUtAeFoSYbKedZNsDvCN on Jupiter
  • Craft the 1000 JUP verification payment transaction for sender 8gMBNeKwXaoNi9bhbVUWFt4Uc5aobL9PeYMXfYDMePE2
  • Submit token metadata update through Jupiter express flow

How to install jupiter-vrfd

1Skills CLI (Vercel)

npx skills add jup-ag/agent-skills --skill jupiter-vrfd -a claude-code

2Manually: put the files in place

~/.claude/skills/jupiter-vrfd/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/jupiter-vrfd/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about jupiter-vrfd

How to install jupiter-vrfd in Claude Code?

Run npx skills add jup-ag/agent-skills --skill jupiter-vrfd -a claude-code.

What does jupiter-vrfd do?

It verifies token eligibility on Jupiter, prepares the 1000 JUP payment transaction, and submits metadata updates via public API endpoints.

What is required to run jupiter-vrfd?

An x-api-key from developers.jup.ag, network access to api.jup.ag, and 1000 JUP to cover the verification fee.

Source text

SKILL.md · MIT
---
name: jupiter-vrfd
description: Use when a user mentions Jupiter token verification, VRFD eligibility, paying 1000 JUP to verify a token, submitting a verification request, or updating metadata via the Jupiter express verification flow.
license: MIT
metadata:
  author: jup-ag
  version: "1.0.1"
tags:
  - jupiter
  - jup-ag
  - jupiter-vrfd
  - token-verification
  - verified
  - solana
---

# Jupiter Token Verification

This skill routes agents through the public Jupiter token-verification flow for a Solana token mint.

**Base URL**: `https://api.jup.ag`
**Auth**: `x-api-key` from [developers.jup.ag](https://developers.jup.ag/) (required)
**Cost**: 1000 JUP

## Use/Do Not Use

Use when:

- checking whether a token is eligible for submission
- crafting and signing the submission payment transaction
- executing the submission flow
- optionally updating token metadata as part of the submission
- submitting a metadata-only paid update when eligibility allows metadata but not verification

Do not use when:

- the agent would need private or internal routes
- the agent needs to fetch or merge existing metadata from non-public endpoints
- the user wants swaps, trading, or unrelated Jupiter flows

**Triggers**: `verify token`, `submit verification`, `check eligibility`, `craft payment transaction`, `execute payment`, `pay for verification`, `update token metadata`, `metadata-only submission`

## Intent Router

| User intent               | Endpoint                                                             | Method |
| ------------------------- | -------------------------------------------------------------------- | ------ |
| Check eligibility         | `/tokens/v2/verify/express/check-eligibility?tokenId={TOKEN_ID}`     | `GET`  |
| Craft payment transaction | `/tokens/v2/verify/express/craft-txn?senderAddress={SENDER_ADDRESS}` | `GET`  |
| Sign and execute payment  | `/tokens/v2/verify/express/execute`                                  | `POST` |

## Eligibility Decision Matrix

| `canVerify` | `canMetadata` | Action                                                            |
| ----------- | ------------- | ----------------------------------------------------------------- |
| `true`      | `true`        | verification+metadata (if user has metadata) or verification only |
| `true`      | `false`       | verification only, omit `tokenMetadata`                           |
| `false`     | `true`        | metadata-only                                                     |
| `false`     | `false`       | **STOP** — show `verificationError` / `metadataError` to user     |

## Examples

Load these on demand:

- **[API Reference](./references/api-reference.md)** for the exact request and response shapes, accepted input formats, normalization rules, submission-mode field requirements, and token metadata fields. This is the source of truth for request construction.
- **[Verify](./examples/verify.md)** when the user wants to execute a request and has confirmed the paying wallet details

## Agent Operating Rules

- Reuse as much as possible from the user's first message. Ask only for missing required fields.
- Never ask the user to paste a raw private key or seed phrase into chat.
- Never print secret values. Only mention non-sensitive file paths, key names, and derived public addresses.
- Do not claim a request was submitted unless you have a real API response or the user explicitly ran the local script themselves.
- If the current agent runtime cannot reach the network, install dependencies, or access local signer files, stop before execution and hand the user the exact local steps instead of fabricating progress.

## Execution Notes

For execute requests in constrained agent environments:

- outbound HTTP and package installation may require approval or user permission
- equivalent shell and package-manager commands are fine; do not block on a specific CLI if the environment already has an equivalent way to run the same steps

## Resources

- **Jupiter Burn Multisig**: `8gMBNeKwXaoNi9bhbVUWFt4Uc5aobL9PeYMXfYDMePE2`
- **JUP Token Mint**: `JUPyiwrYJFskUPiHa7hkeR8VUtAeFoSYbKedZNsDvCN`
- **Jupiter Docs**: [developers.jup.ag](https://developers.jup.ag)
- **Jupiter Verified**: [verified.jup.ag](https://verified.jup.ag)

© jup-ag · MIT · jup-ag/agent-skills

Similar skills

All skills for Finance and business →

From the same set4

Project sponsors

Companies that keep this analytics open