io.github.Wavix/mcp

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

  fail25% of grade

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

  tool "my_numbers_update" param "sms_relay_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "my_numbers_update" param "call_status_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "my_numbers_destinations_update" param "sms_relay_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "speech_analytics_create" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "speech_analytics_create" param "callback_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "speech_analytics_retranscribe" param "callback_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "call_control_create" param "callback_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "sms_and_mms_messages_send" param "callback_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "profile_update" param "sms_relay_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "profile_update" param "dlr_relay_url" description: names an outbound copy/exfil destination (BCC/exfil verb). tool "two_fa_verification_cancel" description: imperative telling the model it must read/send/attach something.

• Supply chain

  fail25% of grade

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

  Python build script executes at build time: network call in build script

• Credential hygiene

  fail15% of grade

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

  tool "sip_trunks_create" has a secret-shaped input parameter "password" — secrets must never be tool parameters (the model would supply/handle the credential). tool "sip_trunks_update" has a secret-shaped input parameter "password" — secrets must never be tool parameters (the model would supply/handle the credential). tool "ten_dlc_brand_vettings_import" has a secret-shaped input parameter "vetting_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.

  Network egress present (pyproject declares an HTTP client dependency) — expected for most servers and not a finding on its own. 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

  unscannable10% of grade

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

  no prior version to diff (first sight of io.github.Wavix/[email protected]); version-drift / rug-pull cannot be evaluated. Cold-start risk is covered by point-in-time checks (injection_surface, supply_chain, credential_hygiene), not here.

• 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)