io.github.oleksiijko/pmb-ai

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

  pass25% of grade

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

  Python build script has no build-time network/exec; repository linkage verified

• Credential hygiene

  fail15% of grade

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

  src/pmb/security/redact.py: embeds a PRIVATE KEY block. tests/integration/test_http_bearer_auth.py: assigns a literal to TOKEN (s3cr…yz (len 23)) — non-provider secret literal; prefer environment intake. tests/security/test_redact.py: embeds a PRIVATE KEY block. (in an example/sample file — lower severity, but a real key here still leaks.) tests/security/test_redact.py: contains a hardcoded JWT (eyJh…5c (len 87)) — in an example/sample file; verify it is not a real key. Reads secrets from the environment (e.g. src/pmb/agent_wrapper/loop.py) — the recommended intake shape.

• Permission scope

  warn15% of grade

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

  references credential paths: src/pmb/cli/commands/manage.py: references a credential path (/.cursor/) references credential paths: src/pmb/cli/connect.py: references a credential path (/mcp.json) references credential paths: src/pmb/cli/doctor.py: references a credential path (/.ssh/) has both shell/exec and network egress — capable of download-and-run, though no such path was observed (src/pmb/agent_wrapper/loop.py: subprocess.* call; src/pmb/agent_wrapper/loop.py: outbound HTTP/socket call). has both shell/exec and destructive filesystem writes (src/pmb/agent_wrapper/loop.py: subprocess.* call; src/pmb/cli/commands/snapshot.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: 15 files (src/pmb/agent_wrapper/loop.py, src/pmb/cli/commands/config.py, src/pmb/cli/commands/daemon.py, +12 more) discloses dynamic code execution: src/pmb/config.py: uses eval()/exec() discloses dynamic code execution: src/pmb/reasoning/images.py: uses eval()/exec() discloses broad/destructive filesystem access: src/pmb/cli/commands/snapshot.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: src/pmb/core/search.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: src/pmb/mcp/daemon.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) 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.oleksiijko/[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

  warn10% of grade

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

  sends telemetry to third-party SDK (Datadog) — disclosure, verify opt-in; manifest/source references analytics behavior — disclosure