Minimal MCP server for DeBridge protocol - enables cross-chain token bridging and transaction management
Server has 6 tools with complete schemas and descriptions. Naming follows verb_noun convention (search_token, get_bridge_quote, create_bridge_order, execute_bridge_transaction, get_supported_chains, check_transaction_status). Descriptions are present and contextual (avg ~150 chars), exceeding the 10-char minimum. However, descriptions lack explicit guidance on when to use each tool vs. alternatives, and error recovery paths are not documented. Parameter descriptions are present but generic in places (e.g., 'Optional slippage tolerance percentage' lacks range constraints). Output schemas are not documented, callers cannot predict response structure. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels (READ_ONLY, WRITE, IRREVERSIBLE). Error handling is minimal: handlers throw generic errors without recovery guidance.
Check the status of a DeBridge transaction and get details about the associated orders.
Create a bridge order to transfer tokens between chains. Use the target asset's full address (e.g., 0xdAC17F958D2ee523a2206206994597C13D831ec7) for dstChainTokenOut, not the ticker (e.g., USDT). EVM to EVM: - Set dstChainTokenOutRecipient to recipient's EVM address - Set dstChainTokenOut to the ERC-20 format address of the token to receive To Solana (7565164): - Set dstChainTokenOutRecipient to Solana recipient address (base58) - Set dstChainTokenOut to the base58 address of the token to receive on Solana From Solana: - Set dstChainTokenOutRecipient to EVM recipient address (0x-prefixed) - Set dstChainTokenOut to the ERC-20 format address of the token to receive
Execute a bridge transaction using tx data from create_bridge_order tool. Always ask for confirmation before proceeding.
Get a quote for bridging tokens between chains. Use search_token first to get correct token addresses.
Get a list of all supported chains with their details including chain IDs, names, and native tokens.
No output schemas documented. Callers cannot predict response structure, forcing LLMs to guess field names and types. This breaks tool chaining (e.g., search_token output must include token addresses for get_bridge_quote input).
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite explicit risk levels. execute_bridge_transaction is IRREVERSIBLE but lacks confirmation/dry-run support. LLMs cannot distinguish safe reads from destructive writes.
Parameter constraints lack specificity. 'slippage' accepts string but no range (0-100) is enforced in schema. 'chainId' accepts any string; valid values should be enumerated or documented with examples. LLMs will hallucinate invalid chain IDs.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
Search for tokens on a specific blockchain. For EVM chains, use 0x-prefixed addresses. For Solana, use base58 addresses.
Error handling is generic. Handlers throw 'Tool {name} failed: {error}' without recovery guidance. No distinction between retryable (network timeout) vs. user-fixable (invalid chain ID) vs. fatal (insufficient balance) errors.
execute_bridge_transaction description says 'Always ask for confirmation before proceeding' but no confirmation mechanism exists. Tool can execute irreversible transactions without user consent. Violates confirmation-request pattern.