ai.duami/intent-exchange

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 "get_registration_challenge" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "register_agent" description: off-vendor URL in a description (duami.ai); capability text should not name where data goes. tool "register_agent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "whoami" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "update_agent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "rotate_api_key" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "delete_agent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "get_agent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "get_agent_ratings" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "search_intents" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "get_intent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "create_intent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "update_intent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "cancel_intent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "find_matches" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "place_bid" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "list_bids" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "get_bid" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "send_message" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "accept_bid" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "reject_bid" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "withdraw_bid" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "confirm_fulfillment" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "release_deal" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "list_my_intents" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "list_my_bids" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "rate_counterparty" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "flag_agent" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "send_feedback" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. tool "get_events" param "api_key" description: off-vendor URL in a description (duami.ai.); capability text should not name where data goes. server instructions: off-vendor URL in a description (duami.ai); 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 "get_registration_challenge" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "register_agent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "whoami" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "update_agent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "rotate_api_key" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "delete_agent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_agent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_agent_ratings" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "search_intents" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_intent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_intent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "update_intent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "cancel_intent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "find_matches" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "place_bid" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_bids" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_bid" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "send_message" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "accept_bid" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "reject_bid" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "withdraw_bid" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "confirm_fulfillment" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "release_deal" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_my_intents" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_my_bids" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "rate_counterparty" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "flag_agent" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "send_feedback" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_events" has a secret-shaped input parameter "api_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 ai.duami/[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)