io.github.subsquid-labs/portal-mcp-server

What we checked

Six static checks, weighted by risk. Every result reflects only what could be observed in the published package and repository — never intent.

• Injection surface

  warn25% of grade

  Tool descriptions/manifest scanned for instruction-injection patterns (imperatives at the model, hidden text, 'ignore previous', data-exfil URLs).

  tool "portal_evm_query_logs" param "addresses" description: bare 'IMPORTANT:' prose in a description (tag form is high-confidence; prose is warn).

• Supply chain

  warn25% of grade

  Package provenance: namespace verification, repo linkage, maintainer count, account age, postinstall scripts, typosquat distance.

  dependency "@modelcontextprotocol/client" is edit-distance 2 from popular package "@aws-sdk/client-s3" (typosquat candidate); dependency "@modelcontextprotocol/node" is edit-distance 2 from popular package "@babel/core" (typosquat candidate); dependency "@modelcontextprotocol/server" is edit-distance 1 from popular package "semver" (typosquat candidate); dependency "hono" is edit-distance 2 from popular package "pino" (typosquat candidate)

• Credential hygiene

  fail15% of grade

  How the server takes secrets (env vs plaintext config vs hardcoded); secrets appearing in tool schemas.

  tool "portal_evm_query_logs" has a secret-shaped input parameter "token_symbols" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_query_transactions" has a secret-shaped input parameter "from_token_symbols" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_query_transactions" has a secret-shaped input parameter "to_token_symbols" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_query_token_transfers" has a secret-shaped input parameter "token_addresses" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_query_token_transfers" has a secret-shaped input parameter "token_symbols" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_get_ohlc" has a secret-shaped input parameter "base_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_get_ohlc" has a secret-shaped input parameter "token0_symbol" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_get_ohlc" has a secret-shaped input parameter "token1_symbol" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_get_ohlc" has a secret-shaped input parameter "token0_address" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_evm_get_ohlc" has a secret-shaped input parameter "token1_address" — secrets must never be tool parameters (the model would supply/handle the credential). tool "portal_hyperliquid_query_fills" has a secret-shaped input parameter "fee_token" — secrets must never be tool parameters (the model would supply/handle the credential).

• Permission scope

  pass15% of grade

  Declared tools vs. breadth (filesystem, network, exec); flags shell-exec and unbounded filesystem access.

  No shell-exec, unbounded/sensitive filesystem, or download-and-run primitives found in scanned source. PASS = no static red flag. Static scanning is blind to code fetched or executed at runtime.

• Version behavior

  warn10% of grade

  Diff of tool definitions between versions; new permissions or changed descriptions in a patch release (the postmark-mcp class).

  vs prior 0.8.2 (patch bump): tool "portal_get_wallet_summary" description changed (wording only, no injection pattern); tool "portal_hyperliquid_get_ohlc" gained input property(ies): from_timestamp, to_timestamp; tool "portal_debug_hyperliquid_query_replica_commands" description changed (wording only, no injection pattern).

• Transport config

  fail10% of grade

  Remote servers: TLS and auth mode (none/token/OAuth). Local servers: whether the manifest indicates it phones home.

  remote MCP endpoint declares NO authentication (unauthenticated remote — nginx-ui/CVE-2026-33032 class)