app.cnvs/whiteboard

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 "open_board" param "url_or_id" description: off-vendor URL in a description (cnvs.app); capability text should not name where data goes. tool "wait_for_update" description: cross-tool redirection ('instead of call…'). server instructions: off-vendor URL in a description (cnvs.app); capability text should not name where data goes.

• Supply chain

  pass25% of grade

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

  repository linkage verified

• Credential hygiene

  fail15% of grade

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

  tool "open_board" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_board" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_preview" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "add_text" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "add_link" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "add_image" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "draw_stroke" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "move" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "erase" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "wait_for_update" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "set_board_mode" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_column" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "update_column" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "delete_column" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_task" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "update_task" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "move_task" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "delete_task" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_tasks" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "set_lane" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "set_column_width" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_tasks" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "query_tasks" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "export_tasks" has a secret-shaped input parameter "access_key" — secrets must never be tool parameters (the model would supply/handle the credential).

• Permission scope

  unscannable15% of grade

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

  No source files, package.json, or pyproject were fetched; cannot infer capability scope.

• 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 app.cnvs/[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)