SolanaTxPlain

Plain-English explanations for every Solana transaction using AI powered co-pilot.

Comment

Inspiration

While exploring Solana transactions during this hackathon, I kept running into the same problem:
even simple transactions are extremely hard to understand unless you already know how Solana works.

Just checking Solscan is a headache if you are not sure what everything says.
Block explorers like Solscan are powerful, but they present raw blockchain data, program names, instruction types, account addresses, etc... that mean very little to everyday users. This not only creates a dangerous gap in knowledge but users are expected to sign transactions they don’t fully understand, especially when interacting with DeFi apps.

After sitting on a pile of Solana signatures, the idea became clear:
What if blockchain transactions could be understood the same way we understand complicated photos or documents - by asking AI to explain what’s happening in plain English? (as if I was 6 years old 😂)

That became the core idea behind SolanaTxPlain.

What the Project Does

SolanaTxPlain converts raw Solana transaction data into "clear, human-readable" explanations.

Given a transaction signature, the system:

  • Fetches on-chain transaction data
  • Identifies the intent (transfer, approval, contract interaction, etc.)
  • Explains what changed in the wallet to the user
  • Highlights fees and programs used clearly
  • Detects potential risk patterns, such as token approvals or unusually large permissions
  • Produces a concise summary that a non-technical user can understand

The goal is not to replace explorers, like Solscan, but to sit besides them as an interpretation layer that explains what just happened and why it matters.

How I Built It

The project is built as an AI-powered analysis pipeline:

  1. Transaction Retrieval
    Transactions are fetched from the Solana blockchain using a transaction signature. This includes instructions, account changes, fees, and programs invoked.

  2. Structured Parsing
    The raw transaction is normalized into a structured format:

    • Wallet balance changes
    • Token balance changes
    • Instruction types (e.g. TransferChecked, ApproveChecked)
    • Programs involved (System Program, SPL Token Program, Compute Budget Program)

  3. AI Reasoning Layer
    An AI model of Gemini interprets the structured data to determine:

    • The high-level intent of the transaction
    • What the user likely meant to do
    • Whether permissions or approvals were granted
    • Whether anything looks risky or unusual

  4. Plain-English Output
    The final output is broken into clear sections for the user to read:

    • One-sentence summary
    • Wallet impact
    • Fees
    • Programs used
    • Risk notes
    • Full AI explanation

This architecture makes the system explainable. Every conclusion is tied back to on-chain facts.

Challenges Faced

The biggest challenge was deciding what matters to users.

On-chain data is extremely verbose, but most of it is irrelevant to someone who just wants to know:

“Did I lose money?”
“Did I give permission to something?”
“Is this risky?”

Another challenge was accurately explaining approval transactions, where no tokens move but future risk is introduced. These are some of the most misunderstood actions in crypto, and required careful wording to avoid false alarms while still warning users appropriately.

Finally, I technically faced small challenges with Gemini credits and reading the Solana data. Learning Solana’s transaction model. especially how programs, instructions, and token accounts interact was a steep but valuable learning curve for me.

What I Learned

With existing experience in fintech industry as a backend engineer, I found myself comfortable working on development in web3.

  • How Solana transactions are structured at a low level compared to others like Ethereum
  • The difference between transfers and permissions (approvals)
  • Why cheap, fast blockchains increase usability and risk
  • How AI can be used not just to generate text, but to reason about structured systems
  • How important explainability is when applying AI to financial systems

Why Solana

Solana is uniquely suited for this project because of its:

  • High transaction throughput
  • Near-zero fees
  • Rich on-chain program interactions

These strengths make Solana ideal for real-world consumer applications, but they also make it easy for users to sign complex transactions without understanding them. SolanaTxPlain directly addresses that gap.

Future Vision

The way I could see SolanaTxPlain evolving is in so many fun ways:

  • A wallet-integrated “Explain this transaction” button
  • A browser extension for transaction previews
  • A security assistant for DeFi interactions
  • An API that wallets and explorers can embed

The long-term vision is simple:
No one should sign a transaction they don’t understand.

Built With

Share this project:

Updates