com.aisenseapi/free-public-tools

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).

  server instructions: off-vendor URL in a description (api.verifyum.com); capability text should not name where data goes. server instructions: off-vendor URL in a description (verifyum.com); 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 "renew_lease" has a secret-shaped input parameter "owner_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "release_lease" has a secret-shaped input parameter "owner_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "complete_lease" has a secret-shaped input parameter "owner_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "read_agent_queue" has a secret-shaped input parameter "read_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "enqueue_agent_queue_job" has a secret-shaped input parameter "write_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "read_agent_queue_job" has a secret-shaped input parameter "read_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "claim_agent_queue_job" has a secret-shaped input parameter "worker_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "ack_agent_queue_job" has a secret-shaped input parameter "worker_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "release_agent_queue_job" has a secret-shaped input parameter "worker_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "renew_agent_queue_job" has a secret-shaped input parameter "worker_token" — 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

  warn10% of grade

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

  vs prior 1.4.0 (minor bump): tool "create_heartbeat" description changed (wording only, no injection pattern); tool "read_agent_inbox" description changed (wording only, no injection pattern); new tool "create_agent_queue" touches a sensitive surface (email/exec/net/fs); new tool "read_agent_queue" touches a sensitive surface (email/exec/net/fs); new tool "enqueue_agent_queue_job" touches a sensitive surface (email/exec/net/fs); new tool "read_agent_queue_job" touches a sensitive surface (email/exec/net/fs); new tool "claim_agent_queue_job" touches a sensitive surface (email/exec/net/fs); new tool "ack_agent_queue_job" appeared; new tool "release_agent_queue_job" appeared; new tool "renew_agent_queue_job" appeared.

• 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)