An AI engineering team that already knows Solana.

The Manager takes a request, splits it into parallel tasks, and hands them to specialists that already know the account model. Then it reviews the work and runs the tests before merging.

Anchor · @solana/kit · LiteSVM · Built on Solana
01/ The crew

Every request becomes parallel work.

The Manager is a persistent lead. It hands the brief to specialists that each hold a different piece of Solana, and reviews what comes back before anything merges. Two agents touching the same program can't clobber each other; each one works in a separate git worktree.

Manager

Splits each request into parallel tasks and assigns the crew.

ProgramRust · Anchor
Frontend@solana/kit · React
TestsAnchor · LiteSVM
ReviewSecurity footguns
IntegrationsSPL · Metaplex · Pyth

The Manager assigns the work. The specialists run in parallel.

02/ The Solana stack

It knows the layers a real project actually touches.

Every agent loads the account model, rent, PDAs and signer checks before it writes a line, plus the token, testing and DeFi layers most newcomers never get to.

CoreAccounts & rent exemption
CorePDAs
CoreCPIs
CoreSigner / owner checks
FrameworkAnchor
TokensSPL Token · Token-2022
NFTMetaplex Core
TestingLiteSVM · Surfpool
Client@solana/kit
DeFiJupiter · Jito
OraclesPyth · Switchboard
WalletWallet Adapter
03/ The reviewer

We planted four bugs. It caught all four.

The Review agent catches the footguns Anchor doesn't catch for you, and explains each one in the review before the merge lands. The same eval runs in CI on every mainline push, so a regression fails the build instead of the audit.

  • Missing signer check

    State mutated for an account that was never required to sign.

  • Non-canonical PDA bump

    A caller-supplied bump accepted where the canonical one is required.

  • Lamport overflow

    Unchecked arithmetic on a balance that wraps in release builds.

  • Token vs Token-2022

    Assumes the legacy token program against a Token-2022 mint.

All four sit on the 24-item footgun checklist the Review agent is graded against in CI.

programs/buggy/src/lib.rsplanted
pub fn withdraw(
    ctx: Context<Withdraw>,
    amount: u64,
    bump: u8,
) -> Result<()> {
    // bump comes from the caller and is
    // never compared to the canonical one
    let seeds = &[
        b"vault",
        ctx.accounts.owner.key.as_ref(),
        &[bump],  // ← not canonical
    ];

    **ctx.accounts.vault
        .to_account_info()
        .try_borrow_mut_lamports()? -= amount;

    Ok(())
}
04/ How it teaches

A tutor that shows its work.

You can read every prompt, rule and knowledge module the system loads, and trace any choice an agent made back to the rule that produced it.

Brain

Prompts you can read

A read-only view of each agent's expertise and the knowledge modules behind it.

Review

Corrected in review

The reviewer explains each footgun where you'll see it, so the lesson lands with the diff.

Merges

A worked example

What lands in the repo is working Solana code you can read and run.

Brain · Solana knowledge, loaded into every agent
RustAnchorPDAs CPIsSPL TokenToken-2022 Metaplex CoreSigner checksOwner checks LiteSVM@solana/kitSurfpool

All of it lives in the repo, and you can read any of it.

05/ Evals

The agents get tested like the code they write.

The planted-bug eval runs in CI on every mainline push and goes red when the reviewer misses a seeded bug. The end-to-end suite drives a real local validator, and both live in the repo for you to run yourself.

planted-bug-review Reviewer catches all four seeded bugs · gates the merge in CI Required
anchor-counter-end-to-end Scaffold, deploy, and increment against real bank state E2E
spl-token-end-to-end Build, deploy, mint and transfer against a local validator E2E
cpi-flow Program A invokes Program B · signer seeds · on-chain log assertions E2E

Built like infrastructure, not a product demo.

The same evals and footgun checks follow each change to the agents. When a Solana standard moves, the prompts catch up and the evals say whether anything broke.

Pre-launch · builds rolling out shortly