Lyris docs

What Lyris is

Lyris is a token due-diligence tool for Robinhood Chain. Paste a token's contract address on the Check page and Lyris reads the chain and DexScreener, scores what it finds by fixed rules, reads the project's own website, compares the site's claims with the chain, and writes a short verdict. Every run ends with a signed receipt.

Next to the token check sits a general agent with live tools, for research and questions about the chain and the wider market.

Read the chain, check the claims, keep the receipt.

Quickstart

1. 1
   Open Check a token

   There is no account to make and nothing to install.

2. 2
   Paste the contract address

   The token's 0x address on Robinhood Chain. Lyris confirms there is an ERC-20 contract there before it starts.

3. 3
   Read the scored report

   Each check shows what was read, its status (good, warn or bad) and the points it earned, grouped into Contract, Holders and Liquidity.

4. 4
   Read the agent's verdict

   The agent reads the project's website and lays its claims next to what the chain shows, then writes red flags, green flags and checks to do before buying.

5. 5
   Keep the receipt

   Copy the signed receipt and check its signature at any time.

What it reads

Everything in the report is read live when you run the check. Results are kept for two minutes, so a second check of the same token in that window returns the same report.

• The contract: name, symbol, decimals, total supply, owner(), the EIP-1967 implementation slot and the bytecode itself
• Every Transfer event since genesis, read in 10M-block windows, up to 80,000 transfers, from which every balance is rebuilt
• The 15 largest holders, each labelled Burn, Pool, Contract, Deployer, From deployer or Wallet
• The deployer: the wallet that sent the token's first mint transaction
• DexScreener: every pool on Robinhood Chain, with liquidity, 24h volume, price, market cap, pool age, website and socials
• For v2 pools: the LP token supply and how much of it sits at burn addresses

Circulating supply is the total supply minus what the dead address holds. Holder percentages are shares of circulating supply.

How each check is scored

Each check earns points against a fixed maximum. The rules are the same for every token.

Score and verdict

The score is the points earned divided by the points possible across the checks that were measured, scaled to 0–100 and rounded.

A check Lyris can't measure for a token is marked unknown and left out of the score, so the score always reflects what was actually read. That covers three cases:

• LP locks on concentrated-liquidity (v3/v4) pools: these positions hold no LP token to read, so the report points you to the team's locker link
• Holder count when the transfer history runs past 80,000 transfers
• Holder checks when no Transfer events come back for the token

Liquidity vs market cap appears when DexScreener lists a market cap, and Pool age appears when it lists a pool creation time. A token with no pool scores zero on Liquidity, Liquidity vs market cap and LP tokens burned.

Site claims vs chain

After the scored report, the agent takes over. If DexScreener lists a website for the token, the agent reads it (and one docs page if the site links one), picks 2 to 4 concrete claims, such as the team, utility, supply, locked liquidity, partners or audits, and compares each one with the report.

The written report then gives a verdict in two or three sentences, red flags and green flags tied to a check or a page, and three practical checks to do yourself before buying. The agent works only from the report and the pages it read.

How a receipt is built

Every agent run and every token check ends with a run receipt. Lyris builds a SHA-256 hash chain: it starts from the hash of your task, folds in each tool step (the tool's name, a hash of its input and a hash of its output), then the hash of the final answer, the model that wrote it and the time. The last hash is the receipt id.

Lyris signs the receipt id with its Ed25519 key. The receipt lists the id, the time, the model, a hash for every step, the answer hash and the signature.

What a receipt proves

• The report is unchanged since the run: change one word of the answer or one tool output and the hashes no longer match
• The report was produced by Lyris: the signature checks out only against Lyris's public key
• Which tools ran, in what order, and which model wrote the answer

A receipt is a faithful record of what Lyris read and wrote. The facts on a project's site remain the project's own claims, and the receipt shows exactly what those pages and data sources said at the time of the run.

Check a receipt

Open the receipt under any finished run and press the check button, or ask the API directly. The public key is published at /api/receipt.

GET  /api/receipt
→ { "alg": "Ed25519", "publicKey": "<base64 SPKI>" }

POST /api/receipt
{ "id": "<receipt id>", "signature": "<signature>" }
→ { "valid": true }

Agent mode and Chat mode

The Agent page has two modes. Agent mode is for work: Lyris plans, calls its tools, and finishes with a Result, its Sources and Next actions, plus a signed receipt. A run uses up to 8 model turns and 14 tool calls. Chat mode answers questions directly.

Your conversations are saved in your own browser.

Tools

The agent's tools read free public data. Each step shows in the chat as it runs, with its source.

Models

Answers come from free models on OpenRouter, and Lyris picks the model for each task: the reasoning model leads for agent runs, token checks, long questions and analytical ones, and the fast model leads for quick questions. When a model is busy, the next one in line takes over. Every answer is labelled with the model that wrote it, and that model is named in the receipt.

Limits

The free models are shared by everyone, so each visitor gets a fair share:

• 10 agent runs an hour, token checks included
• 60 chat messages an hour
• Up to 12,000 characters per message

The token

Lyris's token launches soon on Robinhood Chain. The holder dashboard reads any wallet's token balance straight from the token contract.

The transparency page lists the contract, the liquidity pool and its lock, the team allocation, and which features are live and which are planned.

Planned

Every item in this section is Planned. Each one moves to Live on the transparency page the day it ships.

1. 1
   Planned · Holder credit vouchers

   Every 1M tokens held earns one voucher of agent credit a day, from a 100K minimum, starting at $0.40 a voucher. Claimed from the holder dashboard with one wallet signature.

2. 2
   Planned · Token payment bonus

   Top up agent credit with the token and get 20–25% more credit, with extra holder discounts from 100K, 1M and 10M tokens.

3. 3
   Planned · Buyback and burn

   Tokens paid for credit and a share of revenue go to burns. Every burn will be listed on the holder dashboard with a link to its transaction.

4. 4
   Planned · Sealed runs

   Runs executed inside a hardware enclave, with the receipt signed by a key that lives in the enclave.

5. 5
   Planned · More models

   Larger paid models alongside the free ones, picked per task the same way.