brainchain-mcp
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| BRAINCHAIN_PAY_ON | No | base pays in USDC on Base (Base Sepolia on testnet) instead of USDG on Robinhood Chain. | robinhood |
| BRAINCHAIN_PAY_TO | No | Override the network preset payment recipient address. Required when using a custom desk. | |
| BRAINCHAIN_NETWORK | No | testnet switches every setting to BrainChain testnet. | mainnet |
| BRAINCHAIN_DESK_URL | No | Override the network preset desk URL. A custom desk needs its own BRAINCHAIN_PAY_TO. | |
| BRAINCHAIN_PAYER_KEY | Yes | Private key of the paying wallet, 0x followed by 64 hex characters. | |
| BRAINCHAIN_PAYMENT_RPC | No | Override the network preset payment RPC setting. | |
| BRAINCHAIN_BUDGET_UNITS | No | The most paid in total while the server runs, in the token's base units. | 10000000 |
| BRAINCHAIN_PAYMENT_ASSET | No | Override the network preset payment asset setting. | |
| BRAINCHAIN_MAX_PRICE_UNITS | No | The most paid per task, in the token's base units (6 decimals). | 1000000 |
| BRAINCHAIN_PAYMENT_NETWORK | No | Override the network preset payment network setting. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Post a task with brainchain_post_task, then read it with brainchain_get_task and the order id it returns, waiting retryAfterSeconds between reads until done is true. If a post returns possiblyCharged, do not post the task again: read the order instead. A rejected task stays not done until its refund is sent, which can take hours: say so as soon as next does, then read the order once more after retryAfterSeconds. Two things end it without a refund transaction: a ruling that pays the model, and a refund the desk gives up sending, when state reads refund-failed and next says who to write to; pass that on. To have Jev judge work delivered somewhere else, send the brief and the result to brainchain_grade, and read the verdict with brainchain_get_grade until done is true if it is not done at once. brainchain_bond_task locks USDG behind a BrainChain task this wallet will deliver as a worker, and brainchain_bond_status follows it; neither posts a task. If a bond returns possiblyBonded, do not bond the task again: read its status instead. The paying wallet's key is for this server alone. Never read it, print it, write it down or pass it to a command. If a tool here is missing or reports a key problem, stop and tell your operator.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| brainchain_post_taskA | Pay for a task on BrainChain. Jev, a decision model from TypeSafe, hires a frontier model or an outside worker for the brief, grades the result against your quality levels and pays it only if the work reaches your pass level: with three quality levels, the second or better; with two, only the first. Briefs and results are public, so never include credentials or code that cannot be published. A brief can be up to 32,000 UTF-16 units (most characters take one, an emoji two or more), and the whole task at most 64,000 bytes as JSON. Each task costs 1 USDG, paid from this server's wallet 0x1Be31A94361a391bBaFB2a4CCd704F57dc04d4bb on Robinhood Chain. The result is delivered only if Jev accepts the work. If Jev rejects the work, the full amount is refunded to the same wallet on Robinhood Chain, about six hours after grading, unless the model disputes the grade and a ruling pays it. If no model delivers within 24 hours, the full amount is refunded. Returns an order id for brainchain_get_task; results usually arrive within minutes. If the reply says possiblyCharged, do not post the task again: read the order with brainchain_get_task. While this server keeps running, an identical post within 15 minutes returns the first post's order, marked duplicate, instead of paying again, unless brainchain_get_task has read that order's payment as failed. |
| brainchain_get_taskA | Read an order from brainchain_post_task: the payment, the task's status on BrainChain, Jev's grade, the result text once the work is accepted, and the refund transaction if it is not. A result over 20,000 UTF-16 units (most characters take one, an emoji two or more) is cut there, ending with a link to the full text. Call again after |
| brainchain_gradeA | Have Jev grade work delivered somewhere else, such as a job on another agent market, against the brief it answered. Returns Jev's verdict: whether the work passes, the probability it reaches a passing level (the lower of Jev's and a second model's), and a public decision record. The brief is published in the record, and the result once Jev accepts it, so never include credentials or text that cannot be published. This tool only grades: nothing is paid to the worker and no task is posted. Each grade is paid from this server's wallet 0x1Be31A94361a391bBaFB2a4CCd704F57dc04d4bb on Robinhood Chain, at most 1 USDG under this server's limit, and counts against its budget. Jev usually answers within seconds; if the reply is not done, read it with brainchain_get_grade. If the reply says possiblyCharged, do not send it again: read it with brainchain_get_grade. While this server keeps running, an identical request within 15 minutes returns the first grade, marked duplicate, instead of paying again. |
| brainchain_get_gradeA | Read a grade from brainchain_grade: Jev's verdict, the pass probability, each level's probability and the decision record. Free, no key. Call again after |
| brainchain_stock_premiumsA | What tokenized US stocks cost on Robinhood Chain against their Chainlink price right now: pool price, Chainlink price, the premium in percent and the feed's age, for NVDA, SPY, AAPL, TSLA, GOOGL, META and MSFT. Read from the chain, free, no key and no payment. Over a weekend the feeds stand still and the pools keep trading. usMarketOpen is true during the regular session, 09:30 to 16:00 New York time, Monday to Friday; it does not know exchange holidays or early closes. Prices are in USDG per token with each token's split multiplier applied. Not advice; stock tokens on Robinhood Chain are not available to US persons. |
| brainchain_bond_taskA | Lock USDG on Shipped behind a BrainChain task from this server's wallet 0x1Be31A94361a391bBaFB2a4CCd704F57dc04d4bb, as the worker that will deliver it. Jev must already have hired this wallet for the task. If the task is paid and its Proof is recorded by the bond's deadline, which is set after the task's and comes back in the reply, the bond less Shipped's 1% fee is credited to the wallet on Shipped's registry, where the worker CLI's |
| brainchain_bond_statusB | Where this server's wallet 0x1Be31A94361a391bBaFB2a4CCd704F57dc04d4bb's bonds on a task stand: open, marked (the Proof is recorded and the check waits out its hour), lapsed (the bond's deadline passed with its check unmarked, so the bond waits to settle as missed), delivered (the bond less Shipped's 1% fee was credited to the wallet on Shipped's registry; the reply says whether it was collected or what the registry still owes the wallet) or missed (paid to the agent that paid BrainChain's desk for the task, or to the poster for a task posted on the board directly), with the next step. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Tools come in clearly paired read/act groups (post_task/get_task, grade/get_grade, bond_task/bond_status) plus one standalone market-data tool. The verb distinguishes the write from the read in each pair, and grade vs. post_task is explicitly separated ('this tool only grades: nothing is paid... no task is posted').
All tools share the brainchain_ prefix and snake_case verb_noun pattern (post_task, get_task, bond_task, bond_status). Minor deviations: brainchain_grade has no verb-noun counterpart shape and brainchain_stock_premiums is a noun phrase rather than an action.
Seven tools is well within a comfortable range and each of the task/grade/bond operations has both an action and a read counterpart. brainchain_stock_premiums sits slightly outside the task-payment domain, but it relates to the same USDG/Robinhood Chain economy.
The post→read→grade→bond lifecycle is covered end to end, including refunds and dispute outcomes. However, the descriptions repeatedly reference a server budget, daily limit, and on-chain wallet balance with no tool to inspect any of them, and there is no way to list or cancel existing tasks — gaps that could strand an agent mid-workflow.