io.github.cerebrixos-org/tuning-engines

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

  unscannable25% of grade

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

  No tool descriptions, server instructions, prompts, or resources were fetched; nothing to scan for injection.

• Supply chain

  warn25% of grade

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

  dependency "yaml" is edit-distance 2 from popular package "yarn" (typosquat candidate)

• Credential hygiene

  fail15% of grade

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

  scripts/smoke/inference-smoke.mjs: contains a hardcoded OpenAI API key (sk-t…ey (len 23)). Reads secrets from the environment (e.g. scripts/smoke/inference-smoke.mjs) — the recommended intake shape.

• Permission scope

  fail15% of grade

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

  HIGH: shell/exec capability AND network egress together — download-and-execute risk (scripts/smoke/inference-smoke.mjs: imports child_process and calls exec/spawn; scripts/smoke/inference-smoke.mjs: outbound HTTP/socket call). HIGH: shell/exec capability AND destructive filesystem writes together (scripts/smoke/inference-smoke.mjs: imports child_process and calls exec/spawn; scripts/smoke/inference-smoke.mjs: path traversal ('../') combined with file access). discloses shell/exec capability: scripts/smoke/inference-smoke.mjs: imports child_process and calls exec/spawn discloses broad/destructive filesystem access: scripts/smoke/inference-smoke.mjs: path traversal ('../') combined with file access Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• Version behavior

  fail10% of grade

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

  vs prior 0.4.26 (patch bump): new capability primitive(s) in source: child_process, http-client; new hardcoded outbound host/BCC destination(s): api.tuningengines.com, app.tuningengines.com, example.com.

• Transport config

  warn10% of grade

  Remote servers: TLS and auth mode (none/token/OAuth). Local servers: whether the manifest indicates it phones home.

  manifest/source references analytics behavior — disclosure