io.github.ajc3xc/meridian

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 "request_hitl" description: bare 'IMPORTANT:' prose in a description (tag form is high-confidence; prose is warn). tool "complete_sprint_item" description: imperative telling the model it must read/send/attach something. tool "assign_sprint_waves" description: cross-tool redirection ('instead of call…'). tool "complete_wave_gate" description: imperative telling the model it must read/send/attach something. tool "get_agent_instructions" description: imperative telling the model it must read/send/attach something.

• Supply chain

  fail25% of grade

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

  Python build script executes at build time: network call in build script; package name "@meridianmcp/mcp" is edit-distance 2 from popular package "yup" (typosquat candidate)

• Credential hygiene

  fail15% of grade

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

  tool "accept_handoff" has a secret-shaped input parameter "goal_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "record_handoff_correction" has a secret-shaped input parameter "source_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "resume_wave" has a secret-shaped input parameter "goal_token" — secrets must never be tool parameters (the model would supply/handle the credential).

• Permission scope

  fail15% of grade

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

  HIGH: downloads and pipes straight to a shell (bin/meridian-mcp.js: pipes a download into a shell). discloses shell/exec capability: bin/meridian-mcp.js: imports child_process and calls exec/spawn Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• 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 io.github.ajc3xc/[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

  pass10% of grade

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

  remote endpoint "https://usemeridian.us/mcp/sse" uses TLS (https); remote endpoint uses OAuth (strongest auth mode)