ai.retrodiffusion/pixel-art

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 "authenticate" param "api_key" description: off-vendor URL in a description (retrodiffusion.ai); capability text should not name where data goes. server instructions: imperative telling the model it must read/send/attach something.

• 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 "authenticate" has a secret-shaped input parameter "api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_user_style" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "update_user_style" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "delete_user_style" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_available_styles" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_available_models" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_style_usage" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_balance" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "estimate_inference_cost" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "create_inference" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "list_edit_tools" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "fix_pixel_art" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "run_edit_tool" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "estimate_edit_tool_cost" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "start_inference_job" has a secret-shaped input parameter "rd_api_key" — secrets must never be tool parameters (the model would supply/handle the credential). tool "get_inference_job" has a secret-shaped input parameter "rd_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.retrodiffusion/[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)